Stress is difficult to define. Like the terms motivation and emotion, periodically there are calls to strike stress from the scientific lexicon (e.g., Engle, 1985). Stress variously refers to objective events (stressors), subjective psychological states (being stressed), and physiological responses (e.g., increases in cortisol). Following Selye (1975), in this chapter we refer to the events that precipitate stress reactions as stressors and the responses to those events as stress reactions. Events that have the potential to stimulate stress responses are not stressors for all individuals or at all ages. Intra-individual processes mediate the effect of the event on the response (e.g., Frankenhaeuser, 1979). Stress results when the demands of internal or external events exceed immediately available resources. These demands may be physiological, including being overheated, chilled, and so on. They may also be psychological, including perceived threat, failure of expectation, and social rejection. Such conditions threaten well-being and require a shifting of metabolic resources to fuel the processes needed for self-protection. This shift in metabolic resources favors systems involved in immediate survival and threat-related learning processes. When intense or prolonged, this metabolic shift limits activity in systems performing functions that are future oriented, including functions directed at growth and repair. Shifting resources to maintain organism viability is termed allostasis or stability through change (McEwen, 1998). The capacity to respond to stress through allostatic adjustments is necessary for survival. Increasing evidence suggests that when stress responses are limited or acute, they tend to enhance functioning. However, these adjustments have costs that, if frequent or prolonged, may undermine health and development. Thus, as important as activation is in understanding the psychobiol-ogy of stress, an understanding of the processes that regulate stress reactions is critical.
Two systems orchestrate stress responses in mammals: the L-HPA and the NE-SAM systems (Johnson, Kamilaris, Chrousos, & Gold, 1992). These systems interact in complex ways at all levels of their organization. In the early 1900s Cannon (1936) argued that the SAM system was responsible for coordinating the physiological and behavioral responses necessary to meet external challenges to the constancy of the internal milieu. Building on Bernard's theory that organisms have evolved complex adaptive mechanisms to stabilize their internal states, Cannon proposed the concept of fight/flight to describe the behavioral functions of the SAM system. Later, when Selye (e.g., 1975) presented his theory of the general adaptation syndrome, attention shifted from the SAM to the L-HPA system. Both Cannon and Selye recognized that thoughts and emotions could produce increases in sympathetic and adrenocortical activity even when there were no physical threats to homeostasis. However, it was not until researchers understood that activity of the pituitary gland was under the regulation of hypothalamic releasing and inhibiting factors that the outlines of our current understanding of stress and its relations to the neurobiology of emotion and cognition began to be discerned. It is now well recognized that the SAM and L-HPA system are regulated in part by forebrain structures and pathways, including regions in the prefrontal cortex (Johnson et al., 1992). As in all areas of neuroscience, most of what we know is based on animal research and, when conducted in humans, generally involves adults. Thus, caution is necessary in extrapolating the information presented here to human infants and children.
Contemporary formulations of stress describe a loosely integrated system consisting of neuroanatomical and functional subsystems. Below the neck, stress biology centers on the regulation of glucocorticoids or CORT (cortisol in primates, corticosterone in rodents) and catecholamines, primarily norepinephrine and epinephrine (NE and EPI) (e.g., Johnson et al., 1992). In the periphery, CORT and cate-cholamines operate to increase the energy available for action through inhibiting glucose uptake into storage sites and liberating energy from fat and protein stores. Concurrently, they stimulate increases in cardiovascular and pulmonary function to support the increased motor activity needed in times of challenge. Finally, in concert with central components of the stress system, they function to modulate the biology of growth and repair, including digestion, physical growth, immune function, and reproduction. In the brain, the stress system is orchestrated through reciprocal interactions among NE and hypothalamic and extra-hypothalamic corticotropin-releasing hormone (CRH).
Levels of the stress system mature and become organized over the course of development. In humans, the hypothalamic-brain-stem level develops largely during the prenatal period. Development and integration of limbic and hypothalamic-brain-stem circuits likely occur over the course of infancy (Vazquez, 1998). The frontal cortex is also involved in the regulation of limbic and hypothalamic nuclei. The long period of development of the frontal cortex that extends into adolescence (Huttenlocher, 1994) likely means that a protracted period of development of stress reactivity and regulation in humans exists. A prolonged period of postnatal development of the stress system also suggests that postnatal experience may play critical and multiple roles in emerging individual differences in stress reactivity and regulation (e.g., Heim, Owen, Plotsky, & Nemeroff, 1997). Next we describe each level of the stress system in more detail.
The L-HPA system orchestrates mammalian stress biology through the activity of CRH (e.g., Nemeroff, 1996). CRH is a neuroactive peptide produced in the hypothalamus and in extra-hypothalamic sites. In the hypothalamus its production begins the cascade of events that culminates in increased production of CORT by the adrenal glands. Along with several other secretagogues, CRH regulates the production of adrenocorticotropic hormone (ACTH) by the anterior pituitary (for review, see Palkovits, 1987). Released into general circulation, ACTH binds to receptors on adrenocortical cells in the cortex of the adrenal glands and stimulates the biosynthesis and release of CORT into general circulation. Negative feedback regulates L-HPA activation and CORT production. Current evidence suggests that negative feedback is a widely distributed system involving CORT receptors in, but not limited to, the prefrontal cortex, hypothalamus, hippocampus, and the anterior pituitary gland (e.g., de Kloet, Vreugdenhil, Oitzl, & Joels, 1998; Sanchez, Young, Plotsky, & Insel, 2000).
CRH-producing cells in the hypothalamus receive input from other limbic, hypothalamic, and brain-stem nuclei. As discussed later, NE is a major stimulus of CRH activity in response to psychological stressors. However, multiple neurotransmitter and neuropeptide systems, beyond the NE system, are involved in regulating CRH (Palkovits, 1987). Furthermore, hypothalamic CRH-producing cells also receive input from other nuclei in the hypothalamus, particularly those involved in daily energy flow (Dallman et al., 1993). The net result is that the production of cortisol is not a direct reflection of the individual's emotional state. Rather, it reflects the extent to which signals impinging on the hypothalamus from all sources indicate that extraordinary resources can and must be expended in order to meet the demands of the moment.
Balancing internal and external demands is reflected not only in CRH activity at the level of the hypothalamus, but also in CRH activity at extra-hypothalamic sites (Nemeroff, 1996). CRH is produced in many brain structures that are involved in associating fear and anxiety with activation of the stress system, including the amygdala and prefrontal cortex. In addition, one subtype of the CRH receptor, CRH1, appears specifically to mediate fear-related functions, whereas increasing evidence suggests that CRH2 receptors are more involved in anxiety states (Steckler & Holsboer, 1999). The neuroanatomy of the CRH system has lead to the (likely overly simplistic) view of CRH as the central orchestrator of the stress system, both in terms of endocrine and behavioral responses.
CORT has figured prominently in research on the health consequences of chronic stress. One common fallacy about the L-HPA system is that CORT is necessarily bad for one's health and development. In fact, the relationship between CORT and healthy adaptation is an inverted-U function. Although it appears that chronic or frequent high CORT can be detrimental, it is equally apparent that insufficient CORT
has negative consequences (McEwen, 1998). One hypothesis is that the basis for this inverted- U function lies in the two receptors for CORT, termed mineralocorticoid receptors (MR) and glucocorticoid receptors (GR), and the different functions they mediate (de Kloet et al., 1998). According to this hypothesis, MRs primarily mediate processes that sustain and promote mental and physical health, whereas GRs mediate effects that shunt metabolic resources from growth and repair to catabolic activities needed to manage immediate threats. MRs tend to be occupied when CORT levels are in the basal range. GRs become occupied as CORT levels rise in response to stressors. As GRs become occupied, CRH activity in the hypothalamus is restrained and the stress response is terminated. Activation of this system and activation of GRs is normal and probably has beneficial effects. However, when GRs are occupied chronically, GR-mediated biochemical events can threaten neuronal viability and down-regulate or reduce the GRs available to terminate the stress response, leading to an increase in CORT production. Thus, frequent or prolonged elevations in CORT have been postulated to be one cause of subsequent heightened and prolonged CORT elevations following trauma or chronic adversity. Importantly, early experiences in rodents shape the MR and GR receptor systems (e.g., Caldji et al., 1998; Levine, 1994). Conditions associated with adequate maternal care result in increased MR/GR ratios that allow better containment of the stress response and promotive effects associated with MR occupation to be produced across a wider range of CORT production. Histories of inadequate nurturance result in the opposite pattern of decreased MR/GR ratios.
Although the L-HPA system now figures prominently in research on stress, the older focus on the SAM system has not been lost (see review by Johnson et al., 1992). Consider the catecholamines EPI and NE. EPI is produced by the adrenal medulla and then released into general circulation. EPI acts as a stress hormone, whereas NE produced at synapses is a neurotransmitter. Both EPI and NE act to energize and mobilize the organism for action. Neurons of the hypothalamus and other cell groups within the brain stem are the central coordinators of the sympathetic nervous system (SNS). In the brain, NE-producing neurons originating in the locus coeruleus (LC) project widely throughout the cortex. Although the LC has often been considered a component of central autonomic control, there is little evidence to support this view. LC projections seem to be involved in arousal (Saper, 1995). In addition, LC neurons project to the CRH-producing cells in the hypothalamus, serving as a primary stimulus of increased CRH production and sensitization in response to emotional stressors. In a parallel but independent system, CRH-producing neurons in the amygdala project to the LC, bringing activity of the LC under the regulation of extra-hypothalamic CRH. The central nucleus of the amygdala also stimulates activity of the SAM system via projections to the lateral hypothalamus and brain-stem autonomic nuclei. Although the SAM system has long been associated with stress, its activity is not specific to threatening or aver-sive events. Instead, because of the role of the sympathetic system in supporting rapid energy mobilization, its activity tends to track conditions requiring effort and information processing more generally, rather than those involving distress and uncertainty about outcomes more specifically (e.g., Frankenhaeuser, 1979). Despite this, frequent mobilization of the sympathetic system, particularly in the presence of elevated CORT, can threaten physical health.
The SAM system forms one arm of the autonomic nervous system (ANS). The other arm of this system is the parasym-pathetic nervous system (PNS). Unlike the SAM system, which is sometimes referred to as a diffuse or mass-discharge system, the PNS tends to be more fine-tuned, having discrete effects on the organ systems that it innervates (Hugdahl, 1995). Similar to the health-promotive effects of MRs for the L-HPA system, the PNS primarily promotes anabolic activities concerned with the conservation and restoration of energy (Porges, 1995a, 1995b). The presence of PNS terminals on most organs and tissues innervated by the SAM system allows the PNS to serve as a major regulator of sympathetic effects. Furthermore, although both the PNS and SAM systems have been viewed as efferent systems that carry out work dictated by the brain, both systems also have afferent projections to the brain. These afferent projections not only inform the brain about the status of organs and tissues in the periphery but also allow autonomic regulation of the central nervous system.
Parasympathetic neuronal projections leave the brain through several cranial nerves including the 10th cranial, or vagus nerve, which has been the focus of most of the psy-chophysiological research relating activity of the PNS to stress and emotion (Porges, 1995a, 1995b). In the following description we draw heavily from Porges's work, which has stimulated much of the developmental work on emotion and stress (see also the review by Beauchaine, in press). The primary fibers of the vagus nerve originate in two nuclei in the medulla: the dorsal motor nucleus of the vagus (DMNX), which regulates visceral functions, and the nucleus ambiguus (NA), which regulates functions associated with communication and emotion. In addition, a third medullary nucleus, the nucleus tractus solitarius (NTS) receives many
of the afferent projections traveling through the vagus from peripheral organs. In his polyvagal theory, Porges (1995a) argued that this trinity of nuclei forms the central regulatory component of the vagal system. Efferent projections from the NA, the smart vagus (Vna), are the principal vagal component in vagal cardiac and bronchomotor regulation. The intimate associations between Vna and facial and vocal expressions of emotion, in combination with afferent projections through the NTS, provide pathways through which emotion regulation may contribute to stress regulation, and vice versa. Though still speculative, this polyvagal theory offers a number of insights into the potential role of the PNS in regulating stress biology (Porges, 1995b). Specifically, high-baseline Vna should increase the individual's ability to cope effectively with stress by permitting the lifting of what Porges termed the vagal break, allowing rapid increases in sympathetic activity to shift metabolic resources quickly in response to challenge. In addition, feedback to the NTS via afferent projections of the vagal system should stimulate CNS containment of both the L-HPA and SAM system reactivity.
The physiology of stress can be activated and regulated with little or no input from limbic or cortical centers. Limbic-cortical involvement provides the opportunity to anticipate threats to homeostasis before they are actualized, allowing for preparatory, defensive responses. Integration of corticol-imbic with hypothalamic-brain-stem stress systems also means that feedback and afferent projections of the L-HPA, NE-SAM, and vagal systems influence cognitive-emotional behavior. All attempts to describe the neurobiology of emotion and stress trace their history to work by Papez as elaborated by MacLean (1952). Accordingly, emotions involve the integration of neural structures that include hypothalamic and brain-stem nuclei, along with structures such as the amygdala, hippocampus, cingulate gyrus, and or-bitofrontal cortex (see Figure 5.1).
The amygdala has long been known to mediate adreno-cortical responses to psychosocial stressors (Palkovits, 1987). Its role in negative emotion and conditioned fear is also now well established (for review, see Rosen & Schulkin, 1998). The amygdala and the bed nucleus of the stria terminalis (BNST) form the core structures in current views of the neurobiology of fear, anxiety, and emotional activation of the stress system. The amygdala is comprised of multiple nuclei that are richly interconnected with other parts of the brain. The central nucleus of the amygdala (CEA) has widespread influence over the L-HPA, NE-SAM, and vagal systems via amygdalofugal and stria terminalis pathways. Lesions of the amygdala and surrounding cortex in adult animals prevent elevations in stress hormones to psychological stressors such as physical restraint but do not prevent elevations to physical stressors such as illness or surgery. Such lesions also affect negative emotionality and impair fear conditioning. Although some have speculated that the CEA is involved in anxiety (e.g., with regard to behavioral inhibition, see Kagan, 1994), the role of the CEA in anxiety has recently been questioned. Indeed, Davis has argued that the BNST is more centrally involved in regulating anxious affectivity (for discussion, see Rosen & Schulkin, 1998). Nonetheless, although controversy exists regarding the roles of the CEA and BNST in the regulation of fear versus anxiety, both structures and their circuits are involved in the regulation of L-HPA and SAM system responses to events that elicit negative emotionality.
Current views hold that the threshold for activating the CEA and BNST is regulated by extra-hypothalamic CRH. Similar to stimulation of the CEA, microinfusions of CRH into the CEA produce fear behaviors in primates (reviewed by Rosen & Schulkin, 1998). The fear-inducing effects of CRH are mediated by CRH1 receptors, and experiences that increase fearful reactions to events also tend to increase CRH1 receptors in these regions (for review, see Steckler & Holsboer, 1999). There is also increasing evidence that CRF2 receptors may be involved in regulating anxiety and related states. These facts would seem to argue for a close coupling between fear/anxiety and elevations in CORT. As reflected in syndromes such as posttraumatic stress disorder (PTSD), however, this is not always the case. Whereas elevated NE and EPI have been described in PTSD, remarkably, basal cortisol levels are normal or even suppressed and the L-HPA response to stressors is often dampened although levels of CRH are increased (see review by Yehuda, 1998). Nevertheless, emotion-modulated startle responses, which are believed to reflect responsivity of the CEA and BNST, are increased in animal models of PTSD and are further enhanced by infusions of CRH especially in the presence of high CORT (see review by Rosen & Schulkin, 1998). Odd as it may seem, the limbic CRH and hypothalamic CRH systems appear only loosely coupled. It is not uncommon to find dissociations between these levels of the CRH system and, consequently, between activity of the L-HPA and NE-SAM systems. There is some suggestion that these dissociations may be the result of prolonged elevations in CORT (e.g., Rosen & Schulkin, 1998). In animal models, prolonged CORT elevations produce increased activity of CRH-producing cells in the CEA but decreased activity of similar cells in the hypothalamus. Adrenalectomy (i.e., eliminating CORT) has the opposite effect. Dissociations of this sort may contribute to the development of anxiety disorders (see also Cameron & Nesse, 1988).
Frontal regulation of the limbic, hypothalamic, and brain-stem circuits involved in stress and emotion is a comparatively new frontier in stress research. Although it has long been recognized that the orbitofrontal cortex (OFC) and anterior cingu-late cortex (ACC) play critical roles in regulating emotional behavior (e.g., MacLean, 1952), their roles in regulating activity of the L-HPA and autonomic systems are increasingly appreciated. Indeed, the degree and breadth of interconnectivity between the amygdala and frontal cortex in primates have been one of the surprising findings of the last two decades (Emery & Amaral, 2000). Perhaps especially in primates, the frontal cortex appears to play a central role in stress reactivity and regulation. In this section we briefly describe OFC and ACC regulation of the stress system. Then we broaden the discussion to current views of the roles played by analytic reasoning and positive affectivity.
The OFC and medial cortex have numerous reciprocal connections to the amygdala and other limbic regions (Price, 1999). These connections support the integration of sensory and affective signals, allowing the organization of behavior in relation to reward and punishment. They are also critically important in organizing and modulating behavior so that it is appropriate to the social context. It has been hypothesized that the OFC and its connections to the amygdala and other limbic regions help to mediate attachment effects on stress reactivity and regulation (Schore, 1996). This argument is supported by evidence that the OFC and medial prefrontal regions have connections with hypothalamic and brain-stem regions that regulate behavioral, neuroendocrine, and auto-nomic stress responses. Thus, activity in this region may be important in modulating autonomic and neuroendocrine stress responses.
Technically, the ACC is part of the limbic system. However, it has both cortical and limbic functions and serves, in many ways, to balance activity in the prefrontal regions of the brain with activity in the limbic-hypothalamic areas. The ACC long has been associated with emotion. Most critical to this review, dysregulation of autonomic and neuroendocrine stress reactions are produced by lesions of the ACC (e.g., Diorio, Viau, & Meaney, 1993). The ACC also subserves cognition. It has been hypothesized that the cognitive and emotion functions of the ACC involve two subdivisions, a dorsal cognitive and rostral-ventral affective division (Bush, Luu, & Posner, 2000). According to this perspective, the cognitive division is considered part of the anterior attention network, a distributed attentional network that contributes to executive functioning. The emotional division, on the other hand, is connected to the OFC and medial prefrontal cortex, to the amygdala, and to hypothalamic and brain-stem regions involved in the regulation of stress physiology (e.g., Price, 1999).
Posner and Rothbart (2000) argued that the anterior attention network forms the basis of the effortful control dimension of temperament. Effortful control is believed to contribute importantly to the regulation of social and emotional behavior, particularly when effortful inhibition of actions and emotion are required. Recent evidence that the cognitive and emotional subdivisions of the ACC reciprocally regulate each other may provide one mechanism whereby effortful control exerts inhibitory effects on negative affect and stress physiology (Drevets & Raichle, 1998).
Increases in the size and functional connectivity of the ACC with development may also help explain children's increasing ability to use cognitive coping strategies to regulate emotion, behavior, and stress (e.g., Rothbart, Derryberry, et al., 1994; Wilson & Gottman, 1996).
In addition, affect influences activity of the cognitive and affective subdivisions of the ACC. Positive emotion has been shown to support the cognitive ACC and enhance executive functioning (Ashby, Isen, & Turken, 1999), whereas negative emotion has been shown to decrease activity in the cognitive division (Bush et al., 2000). Thus, conditions that produce anger, fear, and other strong negative affects, if intense, may disrupt children's effortful regulation of their behavior and make it difficult for them to engage in tasks requiring executive function. This ability to dampen negative affects and/or reassert more positive affective states may be critical in regulating stress. Some individuals seem to be able to do this better than others. As discussed in the next section, individual differences may partly reflect asymmetry in neural activity in the prefrontal cortex.
Emotional activity in the prefrontal cortex appears to be lateralized, with activity (for review, see Davidson, 1994; Davidson & Slagter, 2000) in the right prefrontal cortex supporting negative affectivity, while activity in the left supports positive affectivity. It is interesting to note that baseline asymmetry predicts susceptibility to negative and positive emotion-eliciting stimuli and may index the extent of prefrontal-cortex inhibition of limbic-hypothalamic stress circuits. Specifically, greater activity in the right prefrontal cortex may result in disinhibition of the stress system, whereas greater activity in the left prefrontal cortex may help contain and terminate stress reactions. It is not yet clear how this laterality is related to the functioning of specific frontal structures involved in the regulation of the stress response. Nonetheless, the focus on right-frontal asymmetry is consistent with evidence that there is a right bias in the reactive components of the stress system. In rodents there is evidence that the right, not the left, medial frontal cortex mediates neuroendocrine and autonomic responsivity to stressors (Sullivan & Grafton, 1999). Similarly, both sympathetic (Kagan, 1994) and parasympathetic regulation of the heart show a right bias (Porges, 1995a). Hyperactivity in the right frontal regions, then, may reflect a bias not only to negative emotions but also to hyperactivation of the stress system.
Although most of the attention has been on negative emotionality, recently there has been increased attention on positive emotions in stress regulation. Positive affectivity has been associated with problem-focused coping (Folkman & Moskowitz, 2000), perhaps because it supports the engagement of the cognitiveACC and executive functions. Similarly, positive affectivity as reflected in greater left than right frontal activity has been associated with self-reported preferences for approach-oriented coping strategies (Davidson & Irwin, 1999). This is consistent with Davidson's argument that later-alization of emotion in the frontal lobes reflects differential motor biases, with negative emotions organized to support withdrawal and freezing and positive emotions organized to support approach. Greater left than right frontal activity has also been associated with more rapid termination of CEA-generated fear reactions. Davidson and colleagues have suggested that a left-sided bias in the emotion system may allow individuals to experience negative emotions and produce stress reactions to threat, but then to dampen these responses rapidly once the threat has been removed.
The physiology of stress and emotions is complex. While we are beginning to develop a much richer understanding of the neurobiological bases of both emotions and stress, most of the work has yet to be conducted with humans. Furthermore, we know the least about infants and young children. Information about neurobiology, however, can serve as a guide in our attempts to construct a psychobiological account of the development of stress and emotion in early childhood. In addition, the information we are accumulating on young children—when inconsistent with models based on adults or animals—can challenge researchers in neuroscience to provide explanations that are more congruent with the human developmental data. We turn now to what we know about stress and emotions in early human development.
Was this article helpful?