Posttraumatic Stress Disorder

Posttraumatic Stress Disorder

Secondary article Article Contents . Introduction

Robert M Stowe, University of British Columbia, Vancouver, Canada Steven Taylor, University of British Columbia, Vancouver, Canada

. Biological Findings . Cognitive Abnormalities . Neuroimaging Studies

Posttraumatic stress disorder is a clinical condition that develops in some individuals after exposure to major psychological trauma. It is characterized by distressing symptoms, including re-experiencing of the inciting event(s), hypersensitivity to trauma-related stimuli and chronic hyperarousal.

. Summary and Conclusions

Introduction Posttraumatic stress disorder (PTSD) arises after exposure to one or more traumatic events involving death, neardeath or serious injury (or threat thereof). Following the trauma, a constellation of symptoms develops, including (1) re-experiencing of traumatic events, in the form of vivid imagery, thoughts, illusions, hallucinations, eidetic (lifelike) nightmares, and ‘flashbacks’, (2) avoidance of trauma-related thoughts and other stimuli, blunting of emotional responsiveness, estrangement and social withdrawal, and sense of a foreshortened future, and (3) marked arousal or anxiety (insomnia, irritability and temper outbursts, hypervigilance and exaggerated startle responses, and impaired concentration). Diagnosis of PTSD by Diagnostic and Statistical Manual of Mental Disorders version IV (DSM-IV) criteria requires each of the following criteria: (1) a requisite traumatic event, (2) a response characterized by intense fear horror, of helplessness (in children, agitation and disorganized behaviour suffices), and (3) at least one ‘re-experiencing’, three ‘avoidance/numbing’ and two ‘hyperarousal’ symptoms. The disturbance must be disabling and last longer than 1 month. PTSD is a common disorder, with lifetime prevalence estimates in the USA running from 1% to as high as 12.3%. It is associated with alterations in psychophysiological, neurochemical and neuroendocrine functioning, and with structural and functional alterations in certain key areas of the limbic system. Subtle neurological and neuropsychological deficits may occur, and attentional and memory biases for traumatic information have been described. Functional imaging is providing new insights into the operation of distributed neuroanatomical networks concerned with emotional processing, and their dysfunction in PTSD. This chapter provides an overview of these converging lines of evidence.

Biological Findings Neurophysiology and experimental models The psychophysiological paradigm most widely cited as a model of PTSD is fear conditioning, a type of aversive classical conditioning, well studied in rats, cats and other mammals, and in human subjects. Fear conditioning involves pairing of an unconditioned stimulus (UCS) (e.g. foot shock) with a conditioned stimulus (CS), such as a loud noise, a light flash, etc. with the subsequent development of fear to the CS alone. The amygdala, a multinucleated limbic structure with widespread neocortical, limbic and brainstem connections, and a nexus for sensory–limbic integration in the mammalian brain, is thought to be critical in the learning of conditioned responses to fear-inducing stimuli, including learned fear in PTSD. While the central nucleus seems to be the nodal point within the amygdala for visceral and motor effector components of emotional response networks, specific aspects of fear conditioning appear to require contributions from different amygdaloid subnuclei (LeDoux, 1995). The hippocampus, an area critical to the encoding and retrieval of episodic memories; the opioidpeptide rich midbrain periaqueductal grey; and the bed nucleus of the stria terminalis, an area recently implicated in generalized anxiety, appear to play a role in fear conditioned to environmental context. Conditioned fear to explicit cues (as opposed to overall context) does not require the participation of these structures, whereas the amygdala participates in both context-and cue-related types of fear conditioning. Fear conditioning mediated through subcortical sensory input reaching the amygdala directly from subcortical relays can occur with only a single exposure to a fear-eliciting stimulus and is highly resistant to extinction (Le Doux, 1995). Extinction of conditioned fear responses appears to involve inhibition of amygdala

ENCYCLOPEDIA OF LIFE SCIENCES / & 2002 Macmillan Publishers Ltd, Nature Publishing Group / www.els.net

1

Posttraumatic Stress Disorder

activity by medial frontal areas, in particular the orbitofrontal cortex (OFC), a prefrontocortical region with close ties to the limbic and autonomic effector systems, i.e. ‘masking’ rather than ‘erasure’ of amygdala-engrained traumatic memories (Charney et al., 1996). This is consistent with the postulated role of the OFC in reversal of previously learned emotional and stimulus–reward associations and in the inhibition of contextually inappropriate overlearned or ‘automatic’ responses. Acoustic startle reflexes Acoustic startle reflexes have been used to study fear conditioning in both animals and humans. The change in magnitude of the eye-blink response to a noise burst after association of noise with an aversive stimulus is the dependent variable in such paradigms. The reflex can be facilitated not only by prior or simultaneous delivery of an aversive stimulus such as a foot-shock (‘augmented baseline startle’), but also indirectly by the presentation of a cue previously paired with that aversive stimulus (‘fearpotentiated startle’). Fear-potentiated startle is eliminated by lesioning the central or the lateral and basolateral nuclei of the amygdala, or perirhinal cortex, as well as by drugs that block fear and anxiety, and is enhanced by yohimbine, an a2-adrenergic (autoreceptor) antagonist that increases central noradrenergic tone. The acoustic startle reflex involves auditory projections to the amygdala directly from the medial geniculate nucleus, as well as from auditory cortices, and effector projections to the mesencephalic central grey region. It can be inhibited by shortlatency (60–120 ms) warning stimuli delivered before the noise burst, a phenomenon termed pre-pulse inhibition (PPI). A variety of abnormalities of acoustic startle responses have been reported in PTSD of various aetiologies. Reduction of PPI may be a more robust finding than increased startle response amplitudes. The latter have not been found consistently, particularly in unstressed subjects. Whether reduced PPI is a marker for susceptibility or a consequence of stress exposure is uncertain, as it may be a risk factor for psychosis and affective disorder. As a whole, the findings from acoustic startle studies in PTSD appear to support generalization of fear responses both to inappropriate contexts and stimuli (Grillon et al., 1998), consistent with the possibility of decreased prefrontal cortical function and amygdala dysregulation. Event-related potentials Event-related brain potentials (ERPs) are electrical potentials recorded from the scalp (or with intracranial electrodes) using computerized averaging techniques to extract brain activity time-locked to a particular recurring event. ERPs provide an important avenue of research in PTSD because they provide objective data on changes in sensory and attentional processing, including those that 2

occur prior to conscious awareness. Furthermore, error rates on the tasks performed by subjects during collection of ERP data are comparably low in both control and PTSD subjects, implying that electrophysiological differences are more sensitive and less confounded by effort-related factors than task performance measures. ERPs have demonstrated abnormalities in PTSD subjects occurring very early in stimulus processing. For example, when auditory click stimuli are presented to normal subjects, the P50 (sometimes referred to as the P1), a surface-positive potential with a latency of 40–80 ms, is generated. When a second click is delivered about 500 ms later, the P50 to the second stimulus is reduced, a phenomenon referred to as ‘sensory gating’, which may be a correlate of habituation. Decreased sensory gating of the P50 has been found in PTSD (Neylan et al., 1999). This phenomenon may relate to enhanced auditory startle responses in PTSD sufferers. The extensively studied P3 (P300) family of components of the human ERP appears to reflect processes occurring further downstream in the information processing hierarchy, such as stimulus classification, context updating (i.e. comparison of task-relevant stimuli to an internal model or template) and response set (Hillyard et al., 1995). Commonly, the P3 is studied using auditory oddball paradigms, in which subjects have to respond to ‘target’ tones interspersed with more frequent ‘distractor’ tones with a different frequency. ‘P3’, ‘P3a’ and ‘P245’ have been used to denote a frontocentrally dominant positive potential with a latency of around 250–300 ms evoked by infrequent or novel auditory stimuli, whereas ‘P3b’ denotes a parietocentrally dominant P300 waveform that occurs slightly later and is modulated by additional stimulus modalities and conditions such as task relevance (i.e. higher amplitudes to targets than distractors). The amplitude of both the P3a and P3b components is inversely related to stimulus frequency. With at least one notable exception, studies in PTSD have shown reduced P3b amplitude to both targets and distractors, or reduced P3b target effect. These findings are rather nonspecific, may be attributable to co-morbid depression, and may reverse with pharmacological treatment of PTSD (Metzger et al., 1997). When novel (unique) nontarget auditory stimuli are substituted for rare (but repeating) distractors in an oddball task that includes both frequent distractors and rare targets, P3 to these task-irrelevant novels (P3a) is enhanced relative to targets in combat veterans with PTSD but not in those without PTSD, suggesting hyperresponsiveness to novel stimuli in PTSD. Interestingly, this same effect in subjects with orbitofrontal lesions has been interpreted as indexing disinhibition of responses to taskirrelevant stimuli, and may be another physiological correlate of the central disturbances underlying exaggerated startle responses in PTSD. Enhanced amplitude of the mismatch negativity (an ERP component generated to infrequent tones of slightly

ENCYCLOPEDIA OF LIFE SCIENCES / & 2002 Macmillan Publishers Ltd, Nature Publishing Group / www.els.net

Posttraumatic Stress Disorder

deviant pitch in an attended sequence of frequent, identical tones) has been correlated with severity of PTSD symptoms. Changes in the direction and amount of the normal amplitude sensitivity of the P200 component of the auditory ERP evoked by auditory stimuli of differing intensities have been interpreted as reflecting enhanced inhibition of processing of traumatic stimuli in auditory cortex (where both the mismatch negativity and P200 are thought to be generated) by the prefrontal cortex in PTSD. Variable effects on P3 and reaction time latencies have been reported to trauma-related stimuli. P3 amplitudes during an emotional Stroop task (see below) are reduced in PTSD, but do not differ in this regard from patients with other psychiatric disorders. N1, N2 and P2 are early, longlatency components of the ERP which can be elicited even in the absence of conscious attention; however, increased amplitude of waveforms with the same latency are generated by selective attention. Increased amplitude of N1 and P3 to combat-related nontarget pictures, and of N2 to both targets and combat-related pictures, has been reported in PTSD. Amplitude-related augmentation of the auditory P2–N2 components to tone bursts in abused children with PTSD correlated with severity of PTSD symptomatology, particularly flashbacks. Taken together, ERP findings in PTSD are consistent with increased attentional modulation of traumatic stimulus processing, beginning even before conscious awareness can occur; hyperarousal to threat and to novel stimuli; and misallocation of attention to contextually or semantically related but nonthreatening stimuli (i.e. ‘overgeneralization’).

Changes in neuromodulatory systems A great deal of attention in the PTSD neurobiology literature has focused on the noradrenergic system. Combat veterans with PTSD show higher resting heart rate and systolic blood pressure, increased urinary noradrenaline (norepinephrine) excretion, exaggerated cardiovascular responses to combat-related stimuli, panic attacks and hyperarousal, and induction of flashbacks and intrusive traumatic thoughts and emotions by yohimbine. These data are consistent with chronic ‘noradrenergic hyperreactivity’, possibly attributable to a2-adrenergic autoreceptor upregulation in the amygdala and/or to more widespread b-adrenergic receptor hypersensitivity (Charney et al., 1998b). Platelet a2-adrenergics may be reduced in number in PTSD, while existing receptors may be supersensitive to catecholamines (Nutt, 2000). Many symptoms observed in PTSD, including irritability, impulsive aggression, depression and suicidality, and insomnia, have been linked to serotonergic hypofunction in a variety of other clinical and experimental contexts, and various observations suggest that the serotonergic system may be dysfunctional in PTSD. Blunted prolactin release to the serotonergic agonist fenfluramine has been

reported (Davis et al., 1999). Antidepressants that enhance serotonergic tone appear to be the most effective drug treatment for most PTSD symptoms, regardless of whether patients are clinically depressed, and controlled studies have shown efficacy for the selective serotonin-reuptake inhibitors (SSRIs) sertraline and paroxetine, and for tricyclic antidepressants (imipramine and amitriptyline), which inhibit serotonin reuptake nonselectively. The SSRIs appear to improve all three clusters of PTSD symptoms (‘re-experiencing’, ‘avoidance/numbing’ and ‘hyperarousal’) in most studies, although it has been suggested that they are most helpful for avoidant symptoms. Less rigorous studies also suggest that other SSRIs (fluoxetine, fluvoxamine, citalopram) and nefazodone, trazodone, buspirone, lithium, clonazepam and propranolol, all of which have some proserotonergic effects, may be efficacious (Davidson, 2000). Serotonin modulates release of corticotrophin-releasing factor (CRF), and serotonin release during stress may promote glutamatergic excitotoxic injury to hippocampal neurons (Bremner, 1999). Evidence of serotonergic dysfunction is nevertheless nonspecific in PTSD, having been found in many other psychiatric disorders. Prolonged administration of SSRIs results in decreased firing of locus ceruleus neurons, suggesting that SSRIs may improve PTSD symptoms by decreasing noradrenergic hyperactivity. There may be distinct subgroups of PTSD, with some subjects showing heightened sensitivity to serotonergic and others to adrenergic challenges with metachlorophenylpiperazine and yohimbine, respectively (Southwick et al., 1997). A pathophysiological role for other neuromodulators (including dopamine, g-aminobutyric acid and endogenous opioids) has been postulated, but space limitations preclude discussion here.

Neuroendocrine abnormalities While corticosteroid production and secretion increases during acute stress, chronic stress appears to be associated with suppression of cortisol secretion, consistent with a well-replicated finding of low urinary 17-hydroxycorticosteroid excretion and depression of urinary cortisol and noradrenaline (norepinephrine) ratios in PTSD subjects (Charney et al., 1996). Glucocorticoid receptors on lymphocytes are increased in PTSD, consistent with upregulation in response to hypocortisolaemia (Nutt, 2000). Cortisol reduces noradrenergic signal transduction in neurons mediated through the cyclic adenosine monophosphate (cAMP) route, suggesting that chronically low cortisol levels could produce hypersensitivity to adrenergic stimulation. CRF activates the locus ceruleus (thereby increasing noradrenaline release in the neocortex and limbic system), has direct anxiogenic effects and increases startle responses, an effect abolished by lesions of the

ENCYCLOPEDIA OF LIFE SCIENCES / & 2002 Macmillan Publishers Ltd, Nature Publishing Group / www.els.net

3

Posttraumatic Stress Disorder

hippocampus or bed nucleus of the stria terminalis (Charney et al., 1998b). The latter also receives very dense noradrenergic projections, an observation that supports its possible role in mediating or amplifying stress responses (Charney et al., 1998a). An intriguing finding in PTSD is an exaggerated cortisol response (e.g. hypersuppression) to dexamethasone, the converse of the failure of dexamethasone suppression often seen in major depression and some other psychiatric disorders (Charney et al., 1996). This is puzzling because chronic stress in rat models potentiates dendritic atrophy and age-related cell death in the hippocampus (particularly CA3); the accompanying loss of hippocampal glucocorticoid receptors (which normally function in a negative feedback loop regulating the hypothalamic secretion of CRF) results in hypercortisolism and a failure of exogenously administered steroids to inhibit glucocorticoid release (i.e. dexamethasone nonsuppression). A blunted adrenocorticotrophic hormone (ACTH) response to CRF in PTSD would be consistent with a loss of CRF receptors and/or increased negative feedback by circulating glucocorticoids at the pituitary level, as suggested by Charney et al. (1996), but also possibly with a loss of limbic CRF receptors. Hypersecretion of CRF and possibly somatostatin has been reported in Vietnam veterans with PTSD, consistent with both possible aetiologies. Increased consumption of ethanol (which diminishes CRF-evoked ACTH release) by patients with PTSD, however, might also explain this finding. Behavioural sensitization refers to the ability of stressinducing physical stimuli or drugs to produce a state of hyperresponsiveness to subsequent stressors. This process has been studied extensively in rodents after cocaine administration, and reliably induces a cascade of biochemical changes, including rapid induction of expression of the regulatory proto-oncogene protein c-fos in limbic structures, including the hippocampus (Post, 1992). c-fos induction has potentially widespread downstream effects, including modulation of growth factors, neuropeptides, receptors, enzymes and neurotransmitters, and can be produced by activation of noradrenergic, dopaminergic, cholinergic, glutamatergic, opioid and other peptides, as well as by second messengers (e.g. calcium and cAMP). After a single exposure, behavioural sensitization is context dependent and requires an intact amygdala and nucleus accumbens. This selectivity is lost after multiple exposures. Furthermore, a number of lines of evidence suggest that the noradrenergic system may be particularly important in mediating behavioural sensitization, which makes the locus ceruleus hypersensitive to CRF. In these regards, behavioural sensitization may be relevant to PTSD, particularly as a model of the hypersensitivity to environmental stimuli presented in an unrelated context that may be only semantically, rather than physically, related to the originally eliciting traumata (Charney et al., 1998a). 4

Cognitive Abnormalities Deficits in information processing Several studies have examined the relationship between intelligence quotient (IQ) and PTSD severity. Results have been mixed, although more recent, methodologically sound, studies suggest that PTSD severity is negatively correlated with educational attainment and IQ (e.g. Vasterling et al., 1998), even after controlling for the degree of trauma exposure (Macklin et al., 1998). Performance on neuropsychological tests and ‘soft sign’ batteries Some studies have been unable to demonstrate cognitive dysfunction in PTSD. Methodological confounds make these findings difficult to interpret. Better-designed studies have found neuropsychological abnormalities. Deficits in sustained attention have been identified in most studies (e.g. Vasterling et al., 1998). Most studies have observed impairments in various aspects of memory, including acquisition, free recall and sensitivity to retroactive interference. Deficits in set shifting have also been found to discriminate combat veterans with PTSD from those who do not have this disorder. In a recent study, people with chronic PTSD (related to combat or sexual assault) had lower IQs and more neurological soft signs (particularly difficulty drawing two-and three-dimensional figures and difficulty with motor sequencing tasks) compared with combat or assault survivors without PTSD, even when a history of alcoholism or head injury was factored out. These findings appeared to reflect a higher frequency of reported neurodevelopmental problems in the PTSD group. In general, neuropsychological deficits reported in PTSD are mild and may be accounted for primarily by predisposing factors, rather than a consequence of the illness itself. Age, co-morbid mood and anxiety disorders, substance use and head injury are possible confounding variables that have been partially excluded by some, but not all, studies. Autobiographical memory Although people with PTSD tend to show better recall of trauma-related than neutral material (see below), their recollections of traumatic events are often fragmentary and missing in detail, and they may have difficulty recalling the exact sequence of events. People with PTSD, compared with trauma-exposed controls, also have ‘overgeneral’ autobiographical memories, regardless of whether the memories are for traumatic or nontraumatic events; that is, their memories are vague and lacking in detail. For example, when asked to retrieve specific memories in response to emotionally valued cue words (e.g. happy), they recall general categories of memories (‘when I was

ENCYCLOPEDIA OF LIFE SCIENCES / & 2002 Macmillan Publishers Ltd, Nature Publishing Group / www.els.net

Posttraumatic Stress Disorder

fishing’) rather than specific episodes (‘the fishing trip I took during last summer’) (McNally, 1998). Overgeneral memories are correlated with the occurrence of peritraumatic dissociative symptoms (depersonalization, derealization and psychogenic amnesia), which are risk factors for PTSD. Defective recall of autobiographical memories may aggravate attempts to solve problems in everyday life, and may underlie difficulties envisioning the future, a characteristic feature of PTSD (McNally, 1998).

Trauma-specific processing deficits Attentional bias Several experimental studies have investigated whether people with PTSD differ from controls in their attentional biases for trauma-related information. The modified Stroop test is the most widely used paradigm to investigate such processing. The subject is presented with a series of words, each written in a differently coloured ink, and is instructed to name the ink colour as quickly and accurately as possible, while ignoring word meaning. Using this procedure one can assess the degree of ‘interference’ produced by trauma-related words (e.g. rapist) compared with emotionally neutral words (e.g. apple). The degree of interference is reflected in reaction time prolongation for emotional words, and indexes involuntary reallocation of attention to task-irrelevant processes. Studies using the modified Stroop procedure have demonstrated trauma-related interference in various PTSD populations. The degree of interference is correlated with the severity of re-experiencing symptoms but not with numbing or avoidance symptoms. Thus, the extent to which traumatic memories enter working memory appears to be linked to the degree of attentional bias. Cognitive– behavioural therapy of PTSD reduces the degree of Stroop interference.

Memory biases for trauma-related information People with PTSD too easily and involuntarily recall traumatic events. They show enhanced explicit memory for trauma-related material, compared with neutral material (e.g. better recall of trauma-related words, such as ‘combat’ compared with neutral words, such as ‘carrot’), and superior implicit memory for trauma-related over neutral material. These findings may reflect enhanced encoding of threat cues, and may also underlie reexperiencing of traumatic events (a cardinal feature of PTSD), although disruptions in memory access or inhibitory processes also may play a role in re-experiencing symptoms (McNally, 1998). Persistent rumination about the trauma may interfere with the retrieval of nontraumatic material.

Motivated forgetting of traumatic memories It has long been held that trauma victims can unconsciously repress traumatic memories in an effort to block out the overwhelming emotions associated with these events, and that PTSD symptoms do not emerge until these memories eventually emerge into consciousness. The existence of repression and recovered memories has been hotly debated (Loftus, 1993), and currently the evidence is equivocal. Even when people with PTSD are deliberately instructed to forget trauma-related material, they are typically unable to do so (McNally, 1998). In fact, attempts to suppress traumatic memories produce a paradoxical increase in the frequency of intrusive recollections.

Neuroimaging Studies Structural abnormalities Structural imaging studies using magnetic resonance imaging (MRI) have consistently shown smaller hippocampal volumes in people with combat-or sexual assaultrelated PTSD compared with normal and trauma-exposed controls (Bremner, 1999). Smaller hippocampal volume is associated with greater severity of combat stress exposure, dissociative symptoms and PTSD symptoms, and with short-term memory deficits (Bremner et al., 1995). These findings are probably not attributable to confounds such as alcohol abuse, because no consistent differences in other parameters, such as whole brain or temporal lobe volumes, ventricular size, or caudate and amygdala volumes, have been found. Hippocampal ‘volume loss’ has been related to glutamatergic excitotoxicity (Nutt, 2000) and/or to stressrelated phasic increases in glucorticoid levels, which increase vulnerability to excitotoxic and other insults. However, it is not clear that this finding reflects neuronal damage, because glucocorticoids can produce reversible hippocampal volume loss in animal models and also in humans.

Functional imaging Regional patterns of cerebral activity have been studied in PTSD using functional imaging methods such as positron emission tomography (PET), single-photon emission computed tomography (SPECT) and functional magnetic resonance imaging (fMRI). In such studies, subjects are presented with stimuli that may be trauma-related or neutral. For example, a combat veteran might listen to an audiotape of helicopter sounds and gunfire (trauma stimulus) or a control condition consisting of a tape of white noise (neutral stimulus). Subjects typically report intense re-experiencing of symptoms and distress when presented with traumatic stimuli. Regional differences in brain activity for trauma-related versus neutral stimuli are

ENCYCLOPEDIA OF LIFE SCIENCES / & 2002 Macmillan Publishers Ltd, Nature Publishing Group / www.els.net

5

Posttraumatic Stress Disorder

then computed and compared across PTSD and control groups. Differences in sampled populations, small numbers of subjects, and lack of consistency in stimulus presentation paradigms may account for some of the variability in findings across studies. Greater activation of medial frontal cortex is typically seen during presentation of traumatic versus neutral stimuli, and some groups found a super-normal increase in activation in PTSD subjects (e.g. Shin et al., 1999), but this was not replicated by others, who actually reported a decrease in this area. The medial frontal area has been implicated in the processing of affect-related meaning and in inhibition of amygdala activity. Several studies found activation of the gyrus rectus and related medial orbitofrontal cortex. Administration of yohimbine (0.4 mg kg 2 1) produced increased anxiety in patients with PTSD but not in healthy controls, and more PTSD subjects showed a decrease in prefrontal, parietal, temporal and especially orbitofrontal cortex after yohimbine administration, whereas healthy controls showed an increase. A PET study with 15O-labelled water also showed failure of the orbitofrontal cortex to activate normally in PTSD subjects in response to combat-related sites and sounds (Charney et al., 1998a) Some PET studies of cerebral blood flow report reduced activation of the left inferior frontal area during imagery and recall of traumatic personal events in subjects with PTSD compared with controls (Shin et al., 1999). Processing of traumatic (versus neutral) information was associated with lower activation of the middle temporal gyrus in four studies, but not in a fifth. This pattern was greater in PTSD than in control subjects. The middle temporal cortex, in addition to medial prefrontal cortex, plays a role in the extinction of fear. Activity in the insula, an area closely associated with visceral sensation and function and with control of autonomic responses, is greater during the processing of traumatic (versus neutral) stimuli. Results are conflicting regarding the role of other neuroanatomical structures. Some studies have found that, for people with PTSD, exposure to traumatic versus neutral stimuli, or mental imagery but not visual stimuli of traumatic situations, produces greater activity in the anterior cingulate gyrus. Other studies have found that traumatic stimuli produce lower activity in this structure, and there appears to be no difference in activity levels between people with PTSD and those exposed to traumatic stressors who did not develop PTSD (Shin et al., 1999). Inconsistent results also have emerged regarding activation of the amygdala. Some studies found that traumatic (versus neutral) information increases amygdala activation in subjects with PTSD, whereas others found no differences (e.g. Shin et al., 1999). Unilateral left-sided or right-sided activation has been reported. An fMRI study using backward masking techniques to present emotionally charged faces covertly (i.e. subliminally) found enhanced reactivity of both amygdala to fearful versus happy faces in 6

eight patients with PTSD compared with eight combat trauma-exposed controls (Rauch et al., 2000). Failure to find amygdala activation could be due to the use of less distressing stimuli, or to misidentification of portions of the amygdala with adjacent insular cortex due to poor spatial resolution. In summary, findings suggest that distress and reexperiencing symptoms in response to trauma-related stimuli reflect activation of a distributed network of limbic and paralimbic components in the frontal and temporal lobes. The amygdala and the medial frontal cortex may be hyperresponsive in PTSD, and activation of right hemisphere components may predominate (Rauch et al., 1996), consistent with the greater role of the right hemisphere in emotional processing and expression. Activation of many elements of the network may be nonspecific responses to distress, because the patterns seen in PTSD are similar to those found in symptom provocation studies of obsessive– compulsive disorder and specific phobias. A decrease in activation of the left inferior frontal area in response to distressing stimuli may be more specific for PTSD. This region includes Broca’s area, which is involved in phonological assembly and syntactic aspects of language function, and possibly also in verbalization of personal experience (van der Kolk, 1997). This finding may reflect a relative shift from a verbal–intellectual to an experiential mode of awareness in PTSD, as suggested also by studies of auditory probe ERPs showing evidence of greater right hemisphere activation in PTSD subjects during imagery of previous trauma.

Summary and Conclusions Converging evidence from models of learned fear and acoustic startle and behavioural sensitization; psychophysiological, cognitive and neuropsychological paradigms; neurochemical and neuroendocrine investigations; and functional neuroimaging experiments suggest that the neurobiological changes of PTSD involve limbic hyperreactivity to stimuli contextually and/or semantically associated with the inciting traumatic stressors, mediated by long-lasting and relatively resistant stimulus–response associations formed with the critical participation of the amygdala (LeDoux, 1995). Direct projections to the amygdala from brainstem and subcortical areas involved in early, more general, aspects of stimulus processing must be overridden by inhibitory control from medial and orbitofrontal cortical areas and the hippocampus, which are better positioned near the top of the information processing hierarchy to perceptual differentiation of semantically and contextually related but inherently nonthreatening stimuli from fear-conditioned stimuli (Charney et al., 1996). Over-general autobiographic memories may relate to hippocampal changes (McNally,

ENCYCLOPEDIA OF LIFE SCIENCES / & 2002 Macmillan Publishers Ltd, Nature Publishing Group / www.els.net

Posttraumatic Stress Disorder

1998). Premorbid or acquired deficits in cortical inhibition and decreased hippocampal volume (possibly related to stress-induced phasic increases in cortisol release) may be pathogenetically important in the development of PTSD. Hyperarousal and re-experiencing symptoms may be related to noradrenergic hyperactivity and possibly to abnormalities in serotonergic modulation of limbic and catecholaminergic circuits. Threat-focused attentional and explicit and implicit memory biases might follow from these changes. While it remains to be determined whether some of these findings reflect causes, consequences or coeffects in PTSD, it seems clear that the convergence of cognitive neuroscience and neurobiology in the study of this complex disorder has a great deal of promise in furthering our understanding of how personal experience can profoundly shape subsequent emotional responses and behaviour.

References Bremner JD (1999) Does stress damage the brain? Biological Psychiatry 45: 797–805. Bremner JD, Randall PR, Capelli S et al. (1995) Deficits in short-term memory in adult survivors of childhood abuse. Psychiatry Research 59: 97–107. Charney D, Nagy L, Bremner J et al. (1996) Neurobiological mechanisms of human anxiety. In: Fogel B, Schiffer R and Rao S (eds) Neuropsychiatry, pp. 257–286. Baltimore: Williams and Wilkins. Charney DS, Grillon C and Bremner JD (1998a) The neurobiological basis of anxiety and fear: circuits, mechanisms and neurochemical interactions (Part I). Neuroscientist 4: 35–44. Charney DS, Grillon C and Bremner JD (1998b) The neurobiological basis of anxiety and fear: circuits, mechanisms and neurochemical interactions (Part II). Neuroscientist 4: 122–132. Davidson JR (2000) Pharmacotherapy of posttraumatic stress disorder: treatment options, long-term follow-up, and predictors of outcome. Journal of Clinical Psychiatry 61 (supplement 5): 52–56; discussion 57– 59. Davis LL, Clark DM, Kramer GL, Moeller FG and Petty F (1999) Dfenfluramine challenge in posttraumatic stress disorder. Biological Psychiatry 45(7): 928–930 Grillon C, Morgan CA, Davis M and Southwick SM (1998) Effects of experimental context and explicit threat cues on acoustic startle in Vietnam veterans with posttraumatic stress disorder. Biological Psychiatry 44: 1027–1036. Hillyard S, Mangun G, Woldorff M and Luck SJ (1995) Neural systems mediating selective attention. In: Gazzaniga M (ed.) The Cognitive Neurosciences, pp. 665–681. Cambridge, MA: MIT Press. LeDoux JE (1995) In search of an emotional system in the brain: leaping from fear to emotion and consciousness. In: Gazzaniga M (ed.) The Cognitive Neurosciences, pp. 1049–1061. Cambridge, MA: MIT Press. Loftus EF (1993) The reality of repressed memories. American Psychologist 48: 518–537.

Macklin ML, Metzger LJ, Litz BT et al. (1998) Lower pre-combat intelligence is a risk factor for posttraumatic stress disorder. Journal of Consulting and Clinical Psychology 66: 323–326. McNally RJ (1998) Experimental approaches to cognitive abnormality in posttraumatic stress disorder. Clinical Psychology Review 18: 971– 982. Metzger LJ, Orr SP, Lasko NB and Pitman RK (1997) Auditory eventrelated potentials to tone stimuli in combat-related posttraumatic stress disorder. Biological Psychiatry 42: 1006–1015. Neylan TC, Fletcher DJ, Lenoci M et al. (1999) Sensory gating in chronic posttraumatic stress disorder: reduced auditory P50 suppression in combat veterans. Biological Psychiatry 46: 1656–1664. Nutt DJ (2000) The psychobiology of posttraumatic stress disorder. Journal of Clinical Psychiatry 61 (supplement 5): 24–29; discussion 30– 22. Post R (1992) Transduction of psychosocial stress into the neurobiology of recurrent affective disorder. American Journal of Psychiatry 149: 999–1010. Rauch SL, van der Kolk BA, Fisler RE et al. (1996) A symptom provocation study of posttraumatic stress disorder using positron emission tomography and script-driven imagery. Archives of General Psychiatry 53: 380–387. Rauch SL, Whalen PJ, Shin LM et al. (2000) Exaggerated amygdala response to masked facial stimuli in posttraumatic stress disorder: a functional MRI study. Biological Psychiatry 47: 769–776. Shin LM, McNally RJ, Kosslyn SM et al. (1999) Regional cerebral blood flow during script-driven imagery in childhood sexual abuse-related PTSD: a PET investigation. American Journal of Psychiatry 156: 575– 584. Southwick SM, Krystal JH, Bremner JD et al. (1997) Noradrenergic and serotonergic function in posttraumatic stress disorder. Archives of General Psychiatry 54: 749–758. van der Kolk BA (1997) The psychobiology of posttraumatic stress disorder. Journal of Clinical Psychiatry 58 (supplement 9): 16–24. Vasterling JJ, Brailey K, Constans JI and Sutker PB (1998) Attention and memory dysfunction in posttraumatic stress disorder. Neuropsychology 12: 125–133.

Further Reading Davis M (1992) Analysis of aversive memories using the fear-potentiated startle paradigm. In: Squire LR and Butters N (eds) Neuropsychology of Memory, 2nd edn, pp. 470–484. New York: Guilford Press. Morgan CA and Grillon C (1999) Abnormal mismatch negativity in women with sexual assault-related posttraumatic stress disorder. Biological Psychiatry 45: 827–832. Szabo ST and Blier P (2001) Functional and pharmacological characterization of the modulatory role of serotonin on the firing activity of locus coeruleus norepinephrine neurons. Brain Research 922: 9–20. van der Kolk BA, McFarlane AC and Weisaeth L (eds) (1996) Traumatic Stress: The Effects of Overwhelming Experience on Mind, Body, and Society. New York: Guilford Press.

ENCYCLOPEDIA OF LIFE SCIENCES / & 2002 Macmillan Publishers Ltd, Nature Publishing Group / www.els.net

7