Imaging Stress Effects on Memory: A Review of Neuroimaging Studies

Transcription

1 In Review Imaging Stress Effects on Memory: A Review of Neuroimaging Studies Anda H van Stegeren, PhD 1 Objective: To review and give an overview of neuroimaging studies that look at the role of stress (hormones) on memory. Method: An overview will be given of imaging studies that looked at the role of stress (hormones) on memory. Stress is here defined as the acute provocation of the sympathetic adrenal medullar system and the hypothalamic pituitary adrenal axis in experimental designs. Stress hormone levels can be raised endogenously in response to stimulus material (for example, pictures and films) or social evaluative threat situations (stress tasks). Other studies use the exogenous application of drugs to enhance or decrease stress hormone levels. Finally, we review studies on chronic stress and memory. Results: Stress or emotional arousal, leading to increased noradrenaline or cortisol levels, led to better memory performance when it is applied during perception or encoding. Brain regions involved in this process were medial temporal lobe regions such as amygdala and hippocampus, and several parts of the prefrontal cortex (PFC). High stress levels accompanied by high cortisol levels during retrieval led to impaired memory performance. Sex effects on memory as well as lateralization effects on brain activation were found. Conclusion: High stress levels during encoding and consolidation of emotional material involve increased amygdala and hippocampus activation. The role of the anterior cingulate cortex and other parts of the PFC during perception and encoding of arousing material appears to be a modulating one. However, additional research is needed to shed more light on the nature of the brain changes during stress, especially during retrieval. Can J Psychiatry. 2009;54(1): Clinical Implications Emotional and arousing information is better remembered; it evokes a noradrenergic response and activates a network that involves amygdala and hippocampus. Although normal levels of corticosteroid hormones are essential for good cognitive performance, chronic stress levels of glucocorticoids lead to destructive changes in brain and function. Men and women differ in their response to arousing stimuli and stress tasks; a memory-related lateralization in brain activation is found. This could possibly bear significance for the difference in vulnerability to psychopathology in men and women. Limitations Imaging studies on arousal effects and memory have been focusing on a specific a priori hypothesized network of amygdala and hippocampus. The absence of studies, hence limited findings of activation in other brain areas, does not exclude the involvement of other brain areas in stress effects on memory. Very few studies looked at the effects of stress, high endogenous cortisol levels, or glucocorticoid administration during encoding or retrieval within a scanning environment; conclusions are provisional. 16 La Revue canadienne de psychiatrie, vol 54, no 1, janvier 2009

2 Imaging Stress Effects on Memory: A Review of Neuroimaging Studies Key Words: stress, arousal, memory, amygdala, hippocampus, prefrontal cortex, anterior cingulate cortex, neuroimaging, functional magnetic resonance imaging, positron emission tomography Many psychiatric disorders, such as depressive or anxiety disorders, are characterized by a failure of emotion regulation and disturbed emotional memory. Although the exact nature of the cognitive and memory disturbances is not clear yet, the role of stress in psychopathology is evident. The first rapid part of the stress response is orchestrated by the sympathetic adrenal medullary system. Initiated by the hypothalamus, activation of the adrenal medulla results in release of noradrenaline, leading to physical reactions recognized as a fight or flight response. 1,2 Adrenaline and noradrenaline cannot easily pass the blood brain barrier but can stimulate the vagal nerve, causing increased noradrenergic activity in the brain by its action on the locus coeruleus and the nucleus of the solitary tract. These regions stimulate several brain areas, predominantly the amygdala. 3,4 A second slower response is orchestrated by the HPA axis (described in detail in this issue; see Dedovic et al 5 ) at the end of the chain leading to an increase in cortisol level. In the framework of this review, it is relevant to mention that a regulatory network in the brain related to the stress response is formed by structures that are also high in glucocorticoid receptors, namely, amygdala, hippocampus, and PFC. For example, depressed subjects show multiple impairments in attention, working memory, new learning, and executive function in relation to comparison subjects. These deficits have been related to their stress hormone (cortisol) levels and brain volume: depressed subjects had hypercortisolaemia (a 53% increase in area under the curve) and a reduction in right hippocampal volume (a 6% decrease). 6 However, the image is not univocal: other studies did not find hippocampal changes Abbreviations used in this article ACC CBF EMO fmri HPA LOC MIST NEU PET PFC PTSD TSST anterior cingulate cortext cerebral blood flow emotionally arousing functional magnetic resonance imaging hypothalamic pituitary adrenal locus of control Montreal Imaging Stress Task emotionally neutral positron emission tomography prefrontal cortex posttraumatic stress disorder Trier Social Stress Test or smaller hippocampal volumes, but without changes in cortisol levels in depressed patients. 7,8 Cognitive deficits are one of the core symptoms in mood disorders as well as in anxiety disorders such as PTSD. Comparable findings in the latter patient group are memory impairments that may be related to stress-induced hippocampal damage together with changes in cortisol levels. Although most studies find lower baseline (afternoon) cortisol levels in PTSD patients, 9 11 unchanged as well as higher glucocorticoid levels 12 have been found along with smaller hippocampal volumes in PTSD patients that led to a heated debate Findings from animal studies have been extended to patients with PTSD showing smaller hippocampal and ACC volumes, increased amygdala function, and decreased medial prefrontal and ACC function. In addition, patients with PTSD show increased cortisol and noradrenaline responses to stress. 16 But the discussion on the exact role of low cortisol levels as well as increased cortisol and noradrenaline sensitivity to stress in PTSD patients and how these relate to the volumetric and activation changes in the brain is still alive. Normal levels of corticosteroid hormones protect the brain against adverse events and are essential for cognitive performance. However, high levels of centrally acting corticosteroids have been found to be damaging and disruptive to memory formation. Several studies now explain this paradox by appreciating the specific role of the 2 types of steroid receptors (mineralocorticoid and glucocorticoid) in the various stages of information processing and the context in which corticosteroid-receptor activation takes place. Corticosteroid effects on brain tissue as well as cognition can turn from adaptive into maladaptive when actions via both receptor types are imbalanced for a prolonged time To increase our fundamental knowledge, yet also for a better understanding of psychopathology, basic research has been directed at the physiological processes related to emotional memory. Based on animal studies, and later on patient (brain lesion) studies, the path went via pharmacological manipulation of stress levels and its effect on memory to the relatively new imaging methods that now provide the opportunity for a new perspective on brain structure and activation. How is stressful or emotional information attended to, perceived, encoded, consolidated, and retrieved? A solid amount of research has demonstrated that emotional material is remembered better than neutral material. 21,22 A distinction is made between the arousal (ranging from low to high) and valence (from negative to positive) components of the material. Arousal appears to be the core aspect activating the amygdala, whereas valence properties of a stimulus are mainly processed in the prefrontal parts of the brain. 21,23 Elaborating on a wealth of animal studies, 22,24 27 the role of stress hormones such as noradrenaline and cortisol on The Canadian Journal of Psychiatry, Vol 54, No 1, January

3 In Review memory has also been investigated in humans (see van Stegeren, 33 Hurlemann et al, 34 and Wolf 35 for reviews). Patient Studies on Amygdala and Hippocampal Damage The specific role of the amygdala in emotional information processing (without the memory aspect) has been well reviewed 36,37 and with relevance for this paper can be summarized as follows: The amygdala activates during exposure to aversive stimuli from various sensory modalities. Amygdala responses are modulated by the arousal level or current motivational value of stimuli. Activation of the amygdala is associated with many other motor and autonomic functions, as well as cognitive processes including attention and memory. 36 The first evidence for the essential role of the human amygdala in the labelling, perception, and encoding of emotional information in memory came from patient studies Patients with Urbach-Wiethe disease constitute an interesting group, as this disease is characterized by damage confined to the amygdaloid region, while the hippocampal formation and other brain structures are intact. These patients were studied neuropsychologically and with various functional imaging techniques. Their principal bilateral amygdala damage was confirmed. Cognitively, the patients showed little deviation from control subjects, but they differed emotionally as was evident in their judgment of all emotions in facial expressions as well as in remembering negative and positive pictures Several other case studies of subjects with bilateral amygdala damage show that these patients were disproportionately impaired in memory for emotional stimuli, notably highly aversive material. 38,39,44 47 The hippocampus is a prominent brain structure studied widely for its role in cognition and the regulation of the HPA axis. Early evidence for a significant role of the hippocampus in memory processes in humans stems from patient observations 48 of profound impairments of memory function in patients who underwent bilateral removal of the hippocampus and neighbouring brain areas. 49,50 More recent studies show a specific role of the hippocampus for declarative and spatial memory processes The hippocampus is also critically involved in the regulation of the HPA axis driven hormonal response to psychological and physiological challenges. 54 Other Brain Areas An extensive meta-analysis 37 on the functional neuroanatomy of emotion looked at PET and fmri studies that investigated emotional tasks combined with what the authors refer to as cognitive demand. It yielded the observations that the medial PFC appears to have a general role in emotional processing across all categories and domains of interest, while fear specifically engaged the amygdala. Further, induction by emotional recall or imagery, as well as emotional tasks with cognitive demand, recruited the ACC and insula. The communication within the network of amygdala, hippocampus, and the PFC is heavily depending on noradrenergic and glucocorticoid receptors, and is, therefore, logically stress-depending. The hippocampus, involved in the formation of new memories, is also strongly related to amygdala function in emotional memory and both brain structures showed to be noradrenaline-dependent in this function. 55 Noradrenaline has widespread projections throughout the brain; therefore, it is well-suited to orchestrate functions that are related to arousal and stress. noradrenaline can enhance long-term emotional memory consolidation in the amygdala and hippocampus through actions at alpha-1 and beta adrenoceptors. 56 Critical levels of catecholamine stimulation are needed to optimize PFC cognitive function; high levels of catecholamine released during stress may serve to take the PFC off-line to allow faster, more habitual responses mediated by the posterior and subcortical structures to regulate behaviour. 57 Currently, technologies such as PET and fmri scanning have made it possible to identify several brain regions playing a role in emotional information processing. The main part of this review covers studies in control subjects and is directed at the understanding of what happens in the brain when information is encoded under stressful conditions. Stress will be defined for this purpose as the relatively acute reaction on EMO stimuli (such as pictures, films, or sounds) that lead to an activation of the sympathetic adrenal medullary system and (or) HPA axis, accompanied by increased endogenous stress hormone levels. First, fmri studies using mild EMO paradigms, mediated by increased noradrenaline levels, will be discussed. Second, studies that used more stressful material or procedures that also evoke an increase in cortisol levels will be reviewed. Finally, studies will be discussed that used an exogenously applied drug (for example, hydrocortisone) leading to altered stress hormone levels and its effect on memory. A distinction will be made between various aspects of the memory process, as stress hormones such as cortisol have a differential effect on working memory and retrieval (disruption), compared with encoding and consolidation (enhancement). Although fear conditioning can also be regarded as a stressful learning experience, related fmri studies are beyond the scope of this paper and are well-reported earlier Finally, several studies on the effect of chronic elevated cortisol levels on memory and brain volumes and function will be summarized. 18 La Revue canadienne de psychiatrie, vol 54, no 1, janvier 2009

4 Imaging Stress Effects on Memory: A Review of Neuroimaging Studies Imaging Studies on Mild Stress Related to Memory: The Role of Noradrenaline and the Amygdala Typical experimental studies on the effect of EMO material on memory are organized as follows: control subjects are confronted with emotional, compared with neutral, stimuli (pictures of objects or faces, film clips, word lists, or aversive [for example, combat] sounds) during a scanning procedure with PET or fmri. Their memory is tested either immediately, while still in the scanner (working memory), or days or weeks later for long-term episodic memory performance, in- or outside the scanner. The arousal properties of the stimulus material, most of the time, have been shown by peripheral physiological measures in earlier laboratory studies. 28,63 65 With imaging studies, it was now shown that exposure to aversive stimuli in multiple sensory modalities induces activation of the amygdala. Examples include olfactory, gustatory, visual, and auditory modalities. Therefore, the amygdala s responsiveness to aversive stimuli appears to reflect a common, multimodal feature of amygdala coding. 36 An additional question was: How was arousal conceptualized neurobiologically in the brain? In animal studies, noradrenaline played a key role in the activation of the amygdala and improved memory performance related to the stress tasks. Blocking the noradrenergic response, for example, with the betablocker propranolol, effectively blocked the enhancement in memory for emotional material under placebo condition in clinical studies Imaging studies using PET and fmri first showed that amygdala activity is associated with a facilitation of emotional memory. In a PET study, 69 cerebral glucose metabolism in the right amygdaloid complex of human male subjects at encoding was associated with long-term memory for EMO films. Subjects viewed 2 videos consisting either of 12 EMO films or of 12 relatively NEU films, and rated their emotional reaction to each film clip immediately after viewing it. Three weeks after the second session, memory for the videos was assessed in a free recall test. Subjects recalled significantly more EMO than NEU films, and the glucose metabolic rate of the right amygdala while viewing the EMO films was highly correlated with the number of EMO films recalled. One other PET study 70 and one fmri study 71 also reported highly similar findings with correlations between amygdala activation related to emotional stimuli and subsequent memory. In all 3 studies, people who showed strong amygdala activation in response to a set of positive or negative stimuli (relative to other study participants) also showed superior memory for those stimuli (relative to other study participants). Correlations between amygdala activation and subsequent memory were not observed for NEU stimuli. The human amygdala seems to modulate the strength of conscious memory for events according to emotional importance, regardless of whether the emotion is pleasant or aversive. In an additional fmri study, Canli et al 72 were questioning if the correlation between amygdala activation during encoding and subsequent memory, found in the above-mentioned brain imaging studies, was based on individual differences between participants (therefore, their personal evaluation of the stimulus). In a well-designed study, 10 female participants saw neutral and negative slides and rated the emotional intensity of each picture. Thus separate fmri responses in the amygdala for each scene could be related to the participants report of their experience of emotional intensity at study, and to long-term memory for that experience 3 weeks after scanning. The amygdala had the greatest response to scenes rated as most emotionally intense. The degree of activity in the left amygdala during encoding was predictive of subsequent memory only for scenes rated as most emotionally intense. These findings support the view that amygdala activation enhances memory in relation to the emotional intensity of an experience. 72 Amygdala activation was shown to heavily depend on noradrenaline in additional studies 55,73,74 in which betablockers (noradrenergic antagonists) led to a decreased amygdala response to emotional stimuli, associated with poorer memory performance for the presented material. Two fmri studies focused on memory for emotional and perceptual oddballs in lists of words. 55,73 They showed that enhanced memory for the emotional oddball words was strongly coupled to decrements in memory for items preceding these emotional stimuli and that this effect was more pronounced in women. Both memory effects (enhancement, as well as decrements) were reversed by propranolol and abolished in a patient with selective bilateral amygdala damage, suggesting a common neurobiological noradrenergic substrate within the amygdala. In a second comparable study, subjects received either a betablocker (propranolol) or a placebo and were scanned at encoding as well as at retrieval, 8.5 hours later. They found that verbal items that evoke amygdala activation at encoding evoke greater hippocampal responses at retrieval, compared with neutral items. Administration of the betablocker propranolol at encoding abolishes the enhanced amygdala encoding and hippocampal retrieval effects, despite the fact that propranolol is no longer present at retrieval. The central findings of the last study 74 are closely matching the first two studies. 55,73 Male and female control subjects (n = 48) were presented with emotional and neutral pictures during scanning after they had taken either a betablocker (80 mg propranolol) or a placebo (in random order) on 2 consecutive days. Their memory performance was tested in a surprise recognition test 2 weeks later. We found support for The Canadian Journal of Psychiatry, Vol 54, No 1, January

5 In Review Table 1 Overview of changes in activation or volume in brain areas, related to stress tasks, stress hormones, and memory A. Mild stress and memory Brain area activation Materials Hormonal system Amydala Hippocampus PFC Other areas Correlation with memory Imaging type, study Film clips Pictures Words Pictures Noradrenaline Noradrenaline Noradrenaline Noradrenaline Pictures Noradrenaline + R in men + L in women + + Free recall + Recognition task Free recall + Recognition task + Recognition task PET 69 PET 70 fmri 55,73 fmri 74 71, 72, fmri 74, Tasks B. Stress and no memory Hormonal system Amygdala Hippocampus PFC Other areas Imaging type, study MIST Endogenous cortisol Motor and association areas PET and fmri 83 Angular gyrus Arithmetic task Endogenous cortisol R PFC ACC L insula putamen ASL and MRI 83 C. Acute stress Materials and tasks Hormonal system Amygdala Hippocampus PFC Correlation with personality traits Correlation with memory Imaging type, study Pictures Noradrenaline endogenous cortisol Recognition task fmri TSST + Cortisol response Young + Volume +r with SE +r LOC Cued recall MRI 94 + Cortisol response + Cortisol response to awakening Elderly Young Volume + Volume Cued recall before stress MRI 95 MIST + Medial PFC ACC D. Acute stress r with hypothalamus PET and fmri 54 Materials Exogenous cortisol Amygdala Hippocampus PFC Other areas Correlation with memory Imaging type, study Words Hydrocortisol Retrieval PET 77 Words Hydrocortisol Superior frontal gyrus Retrieval fmri 101 ASL = arterial spin labelling; L = left; R = right; r =correlation; SE = self-esteem; + = increase in activation; = decrease in activation 20 La Revue canadienne de psychiatrie, vol 54, no 1, janvier 2009

6 Imaging Stress Effects on Memory: A Review of Neuroimaging Studies the main hypothesis that increasing emotional intensity of the pictures correlated with increased amygdala activation under placebo condition. Application of the betablocker propranolol selectively attenuated this increased amygdala activation during emotional pictures. 74 Their and our results together suggest that human emotional memory is associated with a beta-adrenergic-dependent modulation of the amygdala. We conclude that the encoding of mildly stressful stimuli correlates with amygdala and hippocampus activation and that this process is noradrenergic-dependent (Table 1, part A). Imaging Effects of Stress and Endogenous Cortisol on Memory Whereas memory encoding and consolidation seems to be enhanced by elevated cortisol levels, 24,75,76 memory retrieval appears to be compromised by high cortisol levels. 24,77,81 Importantly, these effects of cortisol during all stages of the memory processes appear to be sex-dependent. 75,80,82 Several experimental tasks and designs have been capable of evoking (endogenous) cortisol responses in healthy subjects. More or less famous procedures are the TSST and Trier Mental Challenge Task, or, in the scanning environment, the MIST 83 (and see review by Dedevic et al 5 in this issue). It consists of a series of computerized mental arithmetic challenges, along with social evaluative threat components. In 2 independent studies using PET and one using fmri, levels of salivary cortisol were significantly increased performing mental arithmetic. They were linked to activation of motor and visual association cortices, as well as activation of brain structures involved in the performance of these tasks (for example, the angular gyrus). A comparable task was imaged with an arterial spin-labelling perfusion MRI that measured CBF changes associated with mild-to-moderate stress induced by a mental arithmetic task with performance monitoring. 84 Stress levels were verified by self-report and salivary cortisol-level measurements. The change in CBF induced by the stress task was positively correlated with subjective stress ratings and activation in the right ventral PFC and in the left insula and putamen area. These areas, together with the ACC, showed sustained activation after task completion in subjects reporting a high stress level during arithmetic tasks. They show that psychological stress induces negative emotion and vigilance and that the right ventral PFC plays a key role in the central stress response. 84 However, both studies picture solely the effect of stress (tasks) on the brain (Table 1, part B) and no memory performance is included. Very few studies looked at the effects of stress or high endogenous cortisol levels on memory within a scanning environment. In humans, the effect of cortisol or acute stress exposure on memory consolidation outside the scanner were found in several laboratory studies, with contrasting results. Social stress-induced cortisol elevation present at the time of learning seems to impair social memory in humans 85 and impairs declarative memory performance, 86 whereas other studies specified that this might be due to the neuroendocrine responsiveness of the participants. Subjects who respond to the (social) stress-induction with a rise in cortisol (high responders) appear to perform better on memory tasks than nonresponders. 87,88 Additionally, this effect appears to be sex-dependent: subjects showing a larger cortisol response recalled less words from a previously learned word list after the TSST than subjects showing only a small cortisol increase. However, this correlation was solely caused by the strong association observed in men, while no association was observed in women. 81 Further, an enhancing effect of glucocorticoids on consolidation of emotional stimuli appears to be critically depending on the presence of noradrenaline, a model heavily supported by animal data. 4,27,89 Some human studies find support for this model, 90,91 in which high cortisol levels evoked by a stress task lead to better memory for emotional stimuli that were accompanied by higher noradrenaline levels. However, all were conducted outside the scanner, with no imaging data available. In a recent fmri study of our laboratory, we found support for this model in humans when we presented our subjects with sets of EMO and NEU pictures and related the brain response to subjects cortisol levels during the scanning procedure. Subjects with higher endogenous cortisol levels had significantly stronger amygdala responses to EMO pictures, compared with participants with lower cortisol levels. 92 Blocking noradrenaline levels in the brain by application of the betablocker propranolol (80 mg) erased this cortisol-level dependent amygdala activation. 92,93 Another group of investigators simultaneously investigated the relations among hippocampal volume, cortisol reactivity, and memory performance, and associated this with the level of self-esteem in several studies. 54,94,95 The idea was that variations in self-esteem and internal LOC the person s perception of being in control of the outcomes predict the neuroendocrine cortisol response to stress. Second, cumulative exposure to high levels of cortisol over the lifetime was known to be related to hippocampal atrophy. In one study, they investigated 16 healthy young and 23 healthy elderly subjects. Structural images were acquired from all subjects and volumetric analyses were performed on medial temporal lobe structures and whole brain grey matter. Self-esteem and internal LOC were significantly correlated with hippocampal volume both in young and in elderly subjects. In the young, the cortisol response to the psychosocial stress task was significantly correlated both with hippocampal volume and with The Canadian Journal of Psychiatry, Vol 54, No 1, January

7 In Review levels of self-esteem and LOC, while in the elderly, these personality traits moderated age-related patterns of cognitive decline, cortisol regulation, and global brain volume decline. 94 In a following study, 95 they tested and challenged the associations among hippocampal volume, cortisol reactivity, and memory performance in healthy young men only, because several of the previous findings related to elderly men and women. 6,96 The results show that larger hippocampal volume was associated with a significantly stronger cortisol increase in response to the TSST and a significantly greater cortisol response to awakening. Declarative memory performance was tested before and after exposure to the TSST. They also report that a larger hippocampal volume is negatively correlated with memory performance on a task presented before stress, but not with memory performance after exposure to the stressor. Young subjects with larger hippocampus volumes performed worse on the memory task presented before the stress procedure. 95 Very recently, they specifically related the intensity of the stress response, measured as cortisol responsivity, to activation changes in the brain. 54 They exposed human subjects to a psychosocial stressor (the MIST) in one PET (n = 10) and one fmri (n = 40) experiment. Interestingly, they observed a profound deactivation of limbic system components, including hippocampus, hypothalamus, medio-orbitofrontal cortex, and ACC, in subjects who reacted to the stressor with a significant increase of the endocrine stress marker cortisol. Further, in the fmri study, the degree of deactivation in the hippocampus was correlated with the release of cortisol in response to the stress task. They conclude that the observed deactivation of limbic system structures suggests elevated activation at rest and during nonstressful situations. The hippocampus is playing an essential role in the regulation of the HPA axis. Based on these findings they propose a model in which the observed reduction in limbic system activity is essential for the initiation of the stress response. 54 Summarizing the cortisol-related effects of stress on memory the conclusion is that there is an intense relation between cortisol level, hippocampus volume, and cognitive performance, but that the direction of this relation is still under debate. Not only is the hippocampus a key structure in the regulation of the HPA axis and cortisol production, but cortisol levels seem to affect hippocampus volume and function reciprocally. Moreover, this relation is affected or moderated by several personality traits such as self-esteem or LOC. Whether the effects on memory are due to neurobiological changes (for example, in transmission) or directly related to volumetric changes remains to be investigated (Table 1, part C). Imaging Studies on Exogenously Manipulated Stress Levels Beneficial effects of acute glucocorticoid treatment or stress exposure on (emotional) memory consolidation have been found, 68,76,97 whereas studies that administered cortisol before (and during) retrieval report a significant decrease in memory performance. 79 Conflicting results of acute cortisol administration on human memory are reported in a meta-analysis. 98 Interestingly, the negative effects of glucocorticoids on memory retrieval are also more prominent for EMO material, 81,99,100 also supporting the noradrenaline cortisol interaction model mentioned above. 24 Only few studies can be found that were directed at imaging the effect of glucocorticoid administration during encoding or retrieval. The first study using PET observed a reduced CBF in the medial temporal lobe after cortisone treatment that was associated with poorer retrieval performance. 77 An event-related fmri study also reported reduced activation in the hippocampus and the superior frontal gyrus during successful memory retrieval when subjects had received cortisol. 101 Both studies suggest that the effects of cortisol on retrieval are at least in part mediated by a disturbance in activation of the medial temporal lobe. Summarizing these findings, cortisol enhances emotional memory consolidation but impairs emotional memory retrieval. Animal and human studies indicate that emotional arousal (hence noradrenaline) is an important prerequisite for the occurrence of these effects. Similar to the effects on consolidation, the amygdala and the hippocampus appear to be the neuroanatomical sites for the negative effects of glucocorticoids on retrieval (Table 1, part D). Imaging Chronic Stress Effects on Memory Several studies showed that chronic exposure to high levels of corticosteroids leads to hippocampal atrophy both in animals 102 and in humans. 102 However, reviewing the supposed stress hippocampus link, 103 it is concluded that this link, although it surely exists, may not be the only one relating stress to the brain. Authors conclude that the link between stress and memory involves various cognitive processing components that are distributed in different regions of the brain, including the hippocampus, but also in cortical areas such as the frontal and prefrontal cortices. In contrast with the small amount of studies on acute stress or exogenously induced high cortisol levels on the brain, numerous studies have been carried out examining the effects of chronic stress on memory and the correlates in the brain, predominantly in patients. 22 La Revue canadienne de psychiatrie, vol 54, no 1, janvier 2009

8 Imaging Stress Effects on Memory: A Review of Neuroimaging Studies Table 2 Volumetric and functional changes in brain areas related to psychopathology and tasks properties Brain areas Amygdala Hippocampus Medial PFC ACC Hormones Psychopathology Volume Activation Volume Activation Volume Activation Volume Activation Noradrenaline Cortisol Depression Right + + PTSD / Brain areas Amygdala Hippocampus Medial PFC ACC Insula Study Volume Activation Volume Activation Volume Activation Volume Activation Noradrenaline Cortisol Arousal (Labar and Cabeza, 21 Kensinger and Corkin 23 ) Valence (Labar and Cabeza, 21 Kensinger and Corkin 23 ) Emotion processing (Phan et al 37 ) Fear (Morris et al, 58 Buchel and Dolan, 59 Buchel et al, 60 Morris et al, 61 Buchel et al 62 ) EMO recall and cognitive tasks = decrease; + = increase; = strong decrease; ++ = strong increase; Patients with chronic hypercortisolaemia, owing to Cushing syndrome, exhibit cognitive dysfunction. 105 Because glucocorticoid excess is associated with hippocampal damage in animals and the hippocampus participates in learning and memory, several studies explored the relations among hippocampal volume, memory dysfunction, and cortisol levels in patients with Cushing syndrome These studies showed an association among reduced hippocampus volume, memory dysfunction, and elevated cortisol levels in patients with Cushing syndrome. 106 Additional studies showed that hippocampal volume increased and restored, at least in part, in such patients after treatment that reduced cortisol to normal concentrations. Following treatment, hippocampal volume increased by up to 10%. 108,110 After cortisol levels decline to normal concentrations, structural volumetric increase in hippocampal volume is then also accompanied by functional improvement in memory performance. 109 Another line of study involves chronic stress related to psychopathology such as in patients with major depressive disorder or PTSD (Table 2). Both disorders are characterized by disturbed HPA axis activity, hence cortisol deregulation, that have profound effects on brain volumes and activation. The structural and functional changes in the human brain related to PTSD are investigated and reviewed extensively Whereas some studies found a reduction in hippocampal volume in adults with PTSD (for a metaanalysis, see Kitayama et al 123 and Smith 124 ), others questioned this and did not find changes between patients with PTSD and matched control subjects. 125 A further refinement of possible mechanisms in veterans with PTSD was investigated in a recent study by Yehuda and colleagues. 126 They found that veterans with PTSD did not differ from those without PTSD in hippocampal volume, but they did show significantly lower urinary cortisol levels and poorer memory performance on 2 memory tests. Further, smaller left hippocampal volumes were observed in veterans who developed PTSD in response to their first reported traumatic exposure, compared with veterans who had first experienced a traumatic event to which they did not develop PTSD, prior to experiencing a subsequent event that led to PTSD. They conclude that although hippocampal volume was not found to differ between subjects with and without PTSD, smaller hippocampal volumes in PTSD may be associated with specific risk and resilience factors. 126 The Canadian Journal of Psychiatry, Vol 54, No 1, January

9 In Review Sex and Lateralization Issues An emerging topic in this field is the attention for possible sex differences as well as lateralization in one of the hemispheres in the processing of emotional information. This is supported by results of imaging studies with PET and fmri reviewed by Wager and colleagues. 127 Several of their reviewed studies show sex differences when imaging brain activity during the perception of EMO stimuli. Reviewing the results with specific focus on the amygdala, they report that density analyses show lateralized clusters with men showing greater peak density in the right amygdala and women in the left amygdala during the encoding of emotional stimuli. Hypothesizing about possible sex and lateralization consequences of this outcome, researchers expanded an earlier study on men with the same number of women. 128 Consistent to their hypothesis, enhanced activity of the right, but not the left, amygdala in men was related to enhanced memory for the previously presented EMO films. Conversely, enhanced activity of the left, but not the right, amygdala in women was related to enhanced memory for the EMO films. These results demonstrated a sex-related lateralization of amygdala involvement in emotionally influenced memory. In line with this, Canli and colleagues 72 expanded their earlier study on women with the imaging of 12 men. They used event-related fmri to assess whether sex differences in memory for emotional stimuli is associated with activation of different neural systems in men and women. Using a very similar design, brain activation in 12 men and 12 women was recorded while they rated again their experience of emotional arousal in response to neutral and emotionally negative pictures. Recognition memory 3 weeks after scanning revealed that highly emotional pictures were remembered best, and remembered better by women than by men. They also found that men and women activated different neural circuits to encode stimuli effectively into memory. They smartly also compared amygdala activation when the analysis was restricted to pictures rated as equally arousing by both groups. They found support for the sex lateralization hypothesis, with men activating significantly more structures than women in a network that included the right amygdala, whereas women activated significantly fewer structures in a network that included the left amygdala for successfully encoded emotional pictures. 129 Several other studies further support these findings on sex-specific lateralization, although the nature and purpose of this lateralization remains unclear. Also, recently one study argued that men and women may not be all that different, but that previous findings are more related to material-specific aspects (words or language related, compared with facial or pictoral information) than sex-specific aspects. 135 Conclusion Recent developments in neuroimaging have made it possible to look at changes in brain volume and activation during stress paradigms. Although normal levels of stress hormones have an important protective function and mainly lead to better cognitive performances, confrontation with extreme stressful events lead to a cascade of events that are accompanied by clinical symptoms, such as memory impairments and (or) a full-blown psychopathological disorder. Imaging studies that examined the effects of relatively mild stressful events, such as arousing stimuli on later memory performance, show an increasing activated amygdala and hippocampus, related to better memory for this EMO material. 55,69,71,74,128 And although the inhibitory and regulating role of the PFC on limbic structures has been investigated relatively often in clinical studies, the effects of stress tasks or cortisol on the PFC in control subjects are not yet clear. Few studies looked at the effects of a stress task in the scanner and related cortisol levels evoked by the task to PFC and ACC activation, apart from the task-related motor and visual cortical areas. 83,84 Studies that related the stress and endogenous cortisol levels to memory performance and brain activation showed that the interaction of noradrenaline and cortisol evokes amygdala activation, related to the increased memory performance. 92,93 Other studies focused more on the relation between cortisol levels and hippocampal volume, and related this to personality aspects such as self-esteem and LOC. Low self-esteem and low internal LOC were clearly associated with smaller hippocampal volume in young and elderly subjects and significantly larger cortisol responses to a stressful situation. 54,94,95 Other personality traits, such as neuroticism and extraversion, have been related to brain (amygdala) activation, for example, during the perception of emotional pictures, but they were not related to memory performance. 136,137 Future studies could elucidate the role and interaction between various personality factors and the way stress affects memory processes in the brain. Finally, imaging studies with exogenously increased cortisol levels by drugs so far point at reduced activation in the medial temporal lobe related to impaired memory retrieval. 77,101 However, much work needs to be done here to further elucidate the role of stress hormones during the various memory stages and the possible differential effect on memory. In conclusion, together with the sex-specific effects related to cortisol responses and its relation with differential memory effects in men and women, imaging stress effects on memory may be an interesting topic for future studies. It may further elucidate sex-specific vulnerabilities in this area and perhaps herewith relate it to differences in prevalence of various forms of psychopathology. 24 La Revue canadienne de psychiatrie, vol 54, no 1, janvier 2009

10 Imaging Stress Effects on Memory: A Review of Neuroimaging Studies Funding and Support Dr van Stegeren is supported by a grant from the Netherlands Organization for Scientific Research (NWO), NWO VENI grant # The Canadian Psychiatric Association proudly supports the In Review series by providing an honorarium to the authors. References 1. Lovallo WR, Thomas TL. Stress hormones in psychophysiological research: emotional, behavioral, and cognitive implications. In: Cacioppo JT, Tassinary LG, Berntson GG, editors. Handbook of psychophysiology. New York (NY): Cambridge University Press; p Mason JW. A review of psychoendocrine research on the sympathetic-adrenal medullary system. Psychosom Med. 1968;30(5 Suppl): McGaugh JL. The amygdala modulates the consolidation of memories of emotionally arousing experiences. Annu Rev Neurosci. 2004(27): Roozendaal B, Hui GK, Hui IR, et al. Basolateral amygdala noradrenergic activity mediates corticosterone-induced enhancement of auditory fear conditioning. Neurobiol Learn Mem Nov;86(3): Epub 2006 Apr Dedovic K, D Aguiar C, Pruessner JC. What stress does to your brain: a review of neuroimaging studies. Can J Psychiatry. 2009;54(1): O Brien JT, Gado MH, Kraemer HC. A longitudinal study of hippocampal volume, cortisol levels, and cognition in older depressed subjects. Am J Psychiatry. 2004;161(11): Sheline YI, Gado MH, Kraemer HC. Untreated depression and hippocampal volume loss. Am J Psychiatry. 2003;160(8): Sheline YI, Mittler BL, Mintun MA. The hippocampus and depression. Eur Psychiatry. 2002;17(Suppl 3): Yehuda R. Linking the neuroendocrinology of post-traumatic stress disorder with recent neuroanatomic findings. Semin Clin Neuropsychiatry. 1999;4(4): Yehuda R. Biology of posttraumatic stress disorder. J Clin Psychiatry. 2000;61(Suppl 7): Yehuda R. Biology of posttraumatic stress disorder. J Clin Psychiatry. 2001;62(Suppl 17): Lindauer RJ, Olff M, van Meijel EP, et al. Cortisol, learning, memory, and attention in relation to smaller hippocampal volume in police officers with posttraumatic stress disorder. Biol Psychiatry. 2006;59(2): Bremner JD. Hypotheses and controversies related to effects of stress on the hippocampus: an argument for stress-induced damage to the hippocampus in patients with posttraumatic stress disorder. Hippocampus. 2001;11(2):75 81; discussion Sapolsky RM. Atrophy of the hippocampus in posttraumatic stress disorder: how and when? Hippocampus. 2001;11(2): Yehuda R. Are glucocortoids responsible for putative hippocampal damage in PTSD? How and when to decide. Hippocampus. 2001;11(2):85 89; discussion Bremner JD. Traumatic stress: effects on the brain. Dialogues Clin Neurosci. 2006;8(4): Joels M, Karst H, Derijk R, et al. The coming out of the brain mineralocorticoid receptor. Trends Neurosci. 2008;31(1): de Kloet ER, Karst H, Joels M. Corticosteroid hormones in the central stress response: quick-and-slow. Front Neuroendocrinol. 2008;29(2): Epub 2007 Oct de Kloet ER, Joels M, Holsboer F. Stress and the brain: from adaptation to disease. Nat Rev Neurosci. 2005;6(6): de Kloet ER, Oitzl MS, Joels M. Stress and cognition: are corticosteroids good or bad guys? Trends Neurosci. 1999;22(10): Labar KS, Cabeza R. Cognitive neuroscience of emotional memory. Nat Rev Neurosci. 2006;7(1): McGaugh JL. Memory a century of consolidation. Science. 2000;287(5451): Kensinger EA, Corkin S. Two routes to emotional memory: distinct neural processes for valence and arousal. Proc Natl Acad Sci USA. 2004;101(9): Roozendaal B. Stress and memory: opposing effects of glucocorticoids on memory consolidation and memory retrieval. Neurobiol Learn Mem. 2002;78(3): McGaugh JL, McIntyre CK, Power AE. Amygdala modulation of memory consolidation: interaction with other brain systems. Neurobiol Learn Mem. 2002;78(3): McGaugh JL. Memory consolidation and the amygdala: a systems perspective. Trends Neurosci. 2002;25(9): Roozendaal B Curt P Richter award. Glucocorticoids and the regulation of memory consolidation. Psychoneuroendocrinology. 2000;25(3): Bradley MM, Greenwald MK, Petry MC, et al. Remembering pictures: pleasure and arousal in memory. J Exp Psychol Learn Mem Cogn,. 1992;18(2): Cahill L, Prins B, Weber M, et al. Beta-adrenergic activation and memory for emotional events. Nature. 1994;371(6499): Cahill L, McGaugh JL. A novel demonstration of enhanced memory associated with emotional arousal. Conscious Cogn. 1995;4(4): Lupien SJ, McEwen BS. The acute effects of corticosteroids on cognition: integration of animal and human model studies. Brain Res Brain Res Rev. 1997;24(1): van Stegeren AH, Everaerd W, Cahill L, et al. Memory for emotional events: differential effects of centrally versus peripherally acting beta-blocking agents. Psychopharmacology (Berl). 1998;138(3 4): van Stegeren AH. The role of the noradrenergic system in emotional memory. Acta Psychol (Amst). 2008;127(3): Epub 2007 Dec Hurlemann R, Hawellek B, Matusch A, et al. Noradrenergic modulation of emotion-induced forgetting and remembering. J Neurosci. 2005;25(27): Wolf OT. The influence of stress hormones on emotional memory: relevance for psychopathology. Acta Psychol (Amst). 2008;127(3): Epub 2007 Sep Zald DH. The human amygdala and the emotional evaluation of sensory stimuli. Brain Res Brain Res Rev. 2003;41(1): Phan KL, Wager T, Taylor SF, et al. Functional neuroanatomy of emotion: a meta-analysis of emotion activation studies in PET and fmri. Neuroimage. 2002;16(2): Adolphs R, Cahill L, Schul R, et al. Impaired declarative memory for emotional material following bilateral amygdala damage in humans. Learn Mem. 1997;4(3): Adolphs R, Tranel D, Hamann S, et al. Recognition of facial emotion in nine individuals with bilateral amygdala damage. Neuropsychologia. 1999;37(10): Anderson AK, Phelps EA. Lesions of the human amygdala impair enhanced perception of emotionally salient events. Nature. 2001;411(6835): Markowitsch HJ, Calabrese P, Wurker M, et al. The amygdala s contribution to memory a study on two patients with Urbach-Wiethe disease. Neuroreport. 1994;5(11): Siebert M, Markowitsch HJ, Bartel P. Amygdala, affect and cognition: evidence from 10 patients with Urbach-Wiethe disease. Brain. 2003;126(Pt 12): Markowitsch HJ. Differential contribution of right and left amygdala to affective information processing. Behav Neurol. 1998;11(4): Cahill L, Babinsky R, Markowitsch HJ, et al. The amygdala and emotional memory. Nature. 1995;377(6547): Hamann SB, Adolphs R. Normal recognition of emotional similarity between facial expressions following bilateral amygdala damage. Neuropsychologia. 1999;37(10): Mori E, Ikeda M, Hirono N, et al. Amygdalar volume and emotional memory in Alzheimer s disease. Am J Psychiatry. 1999;156(2): Adolphs R, Tranel D, Denburg N. Impaired emotional declarative memory following unilateral amygdala damage. Learn Mem. 2000;7(3): Scoville WB, Milner B. Loss of recent memory after bilateral hippocampal lesions. J Neurol Neurosurg Psychiatry. 1957;20(1): Milner B. The memory defect in bilateral hippocampal lesions. Psychiatr Res Rep Am Psychiatr Assoc. 1959;11: Milner B, Penfield W. The effect of hippocampal lesions on recent memory. Trans Am Neurol Assoc ;(80th Meeting): Izquierdo I, Bevilaqua LR, Rossato JI, et al. The connection between the hippocampal and the striatal memory systems of the brain: a review of recent findings. Neurotox Res. 2006;10(2): Squire LR, Stark CE, Clark RE. The medial temporal lobe. Annu Rev Neurosci. 2004;27: Eichenbaum H. The hippocampus and mechanisms of declarative memory. Behav Brain Res. 1999;103(2): Pruessner JC, Dedovic K, Khalili-Mahani N, et al. Deactivation of the limbic system during acute psychosocial stress: evidence from positron emission tomography and functional magnetic resonance imaging studies. Biol Psychiatry. 2008;63(2): Strange BA, Dolan RJ. Beta-adrenergic modulation of emotional memory-evoked human amygdala and hippocampal responses. Proc Natl Acad SciUSA.2004;101(31): Ramos BP, Arnsten AF. Adrenergic pharmacology and cognition: focus on the prefrontal cortex. Pharmacol Ther. 2007;113(3): Arnsten AF. Catecholamine regulation of the prefrontal cortex. J Psychopharmacol. 1997;11(2): Morris JS, Buchel C, Dolan RJ. Parallel neural responses in amygdala subregions and sensory cortex during implicit fear conditioning. Neuroimage. 2001;13(6 Pt 1): Buchel C, Dolan RJ. Classical fear conditioning in functional neuroimaging. Curr Opin Neurobiol. 2000;10(2): Buchel C, Dolan RJ, Armony JL, et al. Amygdala-hippocampal involvement in human aversive trace conditioning revealed through event-related functional magnetic resonance imaging. J Neurosci. 1999;19(24): The Canadian Journal of Psychiatry, Vol 54, No 1, January