Shorter Oxford Textbook of Psychiatry, 6th Ed.

CHAPTER 8. Reactions to stressful experiences


The response to stressful events

Classification of reactions to stressful events

Acute stress reaction and acute stress disorder

Post-traumatic stress disorder

Response to special kinds of severe stress

Adjustment disorders

Special kinds of adjustment


Stressful events frequently provoke psychiatric disorders. Such events can also provoke emotional reactions that are distressing, but not of the nature or severity required for the diagnosis of an anxiety disorder or a mood disorder. These less severe reactions are discussed in this chapter together with post-traumatic stress disorder, which is an intense and prolonged reaction to a severe stressor. With the exception of normal grief reactions, the conditions described in this chapter are listed as disorders in ICD-10 and DSM-IV.

The chapter begins with a description of the various components of the response to stressful events, including coping strategies and mechanisms of defence. The classification of reactions to stressful experience is discussed next. The various syndromes are then described, including acute stress reactions, post-traumatic stress disorder, special forms of response to severe stress, and adjustment disorders. The chapter ends with an account of special forms of adjustment reaction, including adjustment to bereavement (grief) and to terminal illness, and the problems of adults who experienced sexual abuse in childhood.

The response to stressful events

The response to stressful events has three components:

1. an emotional response, with somatic accompaniments

2. a coping strategy

3. a defence mechanism.

Coping strategies and defence mechanisms are overlapping concepts, but they originated from different schools of thought, and for this reason they are described separately in the following account. Individuals have to adapt to stresses, whether they are trying to make sense of specific one-off events, or facing ongoing difficulties that require a change of expectations. The process of coming to terms with stresses is often referred to as ‘working through’ them. This term originated in psychoanalysis to designate the subconscious processing or ‘work’ that was necessary for change. It is now used to mean any coming to terms with emotional difficulties that involves reflection and re-evaluation.

Emotional and somatic responses

These responses are of two kinds:

1anxiety responses, with autonomic arousal leading to apprehension, irritability, tachycardia, increased muscle tension, and dry mouth

2depressive responses, with pessimistic thinking and reduced physical activity.

Anxiety responses are generally associated with events that pose a threat, whereas depressive responses are usually associated with events that involve separation or loss. These features of these responses are similar to, but less intense than, the symptoms of anxiety and depressive disorders (described in Chapters 9 and 10, respectively).

Coping strategies

Coping strategies serve to reduce the impact of stressful events, thus attenuating the emotional and somatic responses and making it more possible to maintain normal performance at the time (although not always in the longer term; see below). The term coping strategy is derived from research in social psychology; it is applied to activities of which the person is aware—for example, deliberately avoiding further stressors. (Responses of which the person is unaware are called defence mechanisms; see below.)

Coping strategies are of two kinds: problem-solving strategies, which can be used to make adverse circumstances less stressful, and emotion-reducing strategies, which alleviate the emotional response to the stressors.

Problem-solving strategies include:

• seeking help from another person

• obtaining information or advice that would help to solve the problem

• solving problems—making and implementing plans to deal with the problem

• confrontation—defending one’s rights, and persuading other people to change their behaviour.

Emotion-reducing strategies include:

• ventilation of emotion—talking to another person and expressing emotion

• evaluation of the problem—to assess what can be changed and try to change it (by problem solving), and what cannot be changed and try to accept it

• positive reappraisal of the problem—recognizing that it has led to some good (e.g. that the loss of a job is an opportunity to find a more satisfying occupation)

• avoidance of the problem—by refusing to think about it, avoiding people who are causing it, or avoiding reminders of it.

Coping strategies are generally useful for reducing the problem or lessening the emotional reaction to it. However, they are not always adaptive. For example, avoidance may not be adaptive in the early stages of physical illness, because it can lead to delay in seeking appropriate treatment. Therefore a person needs not only the ability to use coping strategies, but also the ability to judge which strategy should be used in particular circumstances.

Maladaptive coping strategies

These strategies reduce the emotional response to stressful circumstances in the short term, but lead to greater difficulties in the long term. Maladaptive coping strategies include the following:

• use of alcohol or unprescribed drugs to reduce the emotional response or to reduce awareness of stressful circumstances

• deliberate self-harm either by drug overdose or by self-injury. Some people gain relief from tension by cutting their skin with a sharp instrument to induce pain and draw blood. Others take overdoses to withdraw from the situation or to show their need for help

• unrestrained display of feelings can reduce tension, and in some societies such behaviour is sanctioned in particular circumstances (e.g. grieving). In other circumstances, such behaviour can damage relationships with people who would otherwise have been supportive

• aggressive behaviour—aggression provides immediate release of feelings of anger. In the longer term, it may increase the person’s difficulties by damaging relationships.

Coping styles

When particular coping mechanisms are used repeatedly by the same person in different situations, they are said to constitute a coping style. Some people change their coping strategies according to the circumstances—for example, they may use problem-solving strategies at work but employ avoidance when unwell. Some people habitually use maladaptive strategies—for example, they repeatedly abuse alcohol or take overdoses of drugs when under stress. More recent research has distinguished between coping style, which is seen as a relatively enduring behavioural trait, and coping response, which is much more specific to particular stressful environments.

Defence mechanisms

Defence mechanisms (see Box 8.1) are unconscious responses to external stressors as well as to anxiety arising from internal conflict. They were originally described by Sigmund Freud and later elaborated by his daughter, Anna Freud (1936). In response to stressful circumstances, the most frequent defence mechanisms are repression, denial, displacement, projection, and regression. Defence mechanisms are unconscious processes (i.e. people do not use them deliberately and are unaware of their own real motives, although they may become aware of these later through introspection or through another person’s comments). Freud identified defence mechanisms in his study of the ‘psychopathology of everyday life’, a term that he applied to slips of the tongue and lapses of memory. The concept of defence mechanisms has proved useful in understanding many aspects of the day-to-day behaviour of people under stress, notably those with physical or psychiatric illness. Freud also used the concept of mechanisms of defence to explain the aetiology of mental disorders, but this extension of his original observations has not proved useful.

The main mechanisms of defence are described in Box 8.1

Box 8.1 Defence mechanisms


This is the exclusion from consciousness of impulses, emotions, or memories that would otherwise cause distress. For example, especially painful aspects of the memory of distressing events such as sexual abuse in childhood may be kept out of full awareness for many years.


This is a related concept, which is inferred when a person behaves as if they are unaware of something that they may reasonably be expected to know. For example, on learning that they are dying of cancer, a patient may continue to live normally as if they are unaware of the diagnosis. In this example, denial is adaptive, as it can help to reduce depression. However, in the early stage of illness, denial may delay help seeking or lead to refusal of necessary investigations and treatment. In this second example, denial is maladaptive.


This is the transfer of emotion from a person, object, or situation with which it is properly associated, to another source. For example, after the recent death of his wife, a man may blame the doctor for failure to provide adequate care, and may thus avoid blaming himself for putting his work before his wife’s needs during the last months of her life.


This is the attribution to another person of thoughts or feelings similar to one’s own, thereby rendering one’s own thoughts or feelings more acceptable. For example, a person who dislikes a colleague may attribute reciprocal feelings of dislike to him; it is then easier to justify his own feelings of dislike for the colleague.


This is the adoption of behaviour appropriate to an earlier stage of development—for example, dependence on others. Regression often occurs among physically ill people. In the acute stages of illness it can be adaptive, enabling the person to acquiesce passively to intensive medical and nursing care. If regression persists into the stage of recovery and rehabilitation, it can be maladaptive because it reduces the patient’s ability to make efforts to help him-or herself.

Reaction formation

This is the unconscious adoption of behaviour which is the opposite to that which would reflect the person’s true feelings and intentions. For example, excessively prudish attitudes to sex are sometimes (but not always) a reaction to the person’s own sexual urges that they cannot accept.


This is the unconscious provision of a false but acceptable explanation for behaviour that has a less acceptable origin. For example, a husband may leave his wife at home because he does not enjoy her company, but he may reassure himself falsely that she is shy and would not enjoy going out.


This is the unconscious diversion of unacceptable impulses into more acceptable outlets—for example, turning the need to dominate others into the organization of good works for charity. (There are, of course, many other motives for charitable work.)


This is the unconscious adoption of the characteristics or activities of another person, often to reduce the pain of separation or loss. For example, a widow may undertake the same voluntary work that her husband used to do.

Present circumstances, previous experience, and response to stressful events

Brown and Harris (1978) showed that the response to a stressful life event is modified by present circumstances and by past experience. Some current circumstances make a person more vulnerable to stressful life events—for example, the lack of a confidant with whom to share problems. Such circumstances are called vulnerability factors. Previous experience can also increase vulnerability. For example, the experience of losing a parent in childhood may make a person more vulnerable in adult life to stressful events involving loss. It is difficult to examine these more remote associations scientifically.

Classification of reactions to stressful events

Although they are included within the classifications of diseases, not all reactions to stressful events are abnormal. Grief is a normal reaction to the stressful experience of bereavement, and only a minority of people have a very severe or abnormally prolonged reaction. There is also a normal pattern of reaction to a dangerous or traumatic event such as a car accident. Most people have an immediate feeling of great anxiety, are dazed and restless for a few hours afterwards, and then recover; a few people have more severe and prolonged symptoms—an abnormal reaction. It is difficult to decide where to draw the separation between normal and abnormal reactions to stressful events in terms of severity or duration, and in practice the division is arbitrary. Similarly, among patients who are in hospital for medical or surgical treatment, most are anxious but a few are severely anxious and show extreme denial or other defence mechanisms that impair cooperation with treatment.

ICD-10 and DSM-IV reactions to stressful experiences are classified into three groups (see Table 8.1).

Acute reactions to stress

This category is for immediate and brief responses to sudden intense stressors in a person who does not have another psychiatric disorder at the time. The ICD-10 definition of acute stress reaction requires that the response should start within 1 hour of exposure to the stressor, and that it begins to diminish after not more than 48 hours. The DSM-IV definition of acute stress disorder states that the onset should occur while or after experiencing the distressing event, and requires that the condition lasts for at least 2 days and for no more than 4 weeks. These two definitions capture different phases of the anxiety response as the different terms, reaction and disorder, suggest. ICD-10 refers to the short-lived normal response, whereas DSM-IV refers to the more prolonged response, which is less common. Both diagnostic systems require that the stressor must be of an exceptional nature and, in the case of DSM-IV, that actual or threatened injury to self or others has occurred.

Table 8.1 Classification of reactions to stressful experience


The order of the subgroups has been changed to show the similarities and differences between the two systems.

Post-traumatic stress disorder

This is a prolonged and abnormal response to exceptionally intense stressful circumstances such as a natural disaster or a sexual or other physical assault.

Adjustment disorder

This is a more gradual and prolonged response to stressful changes in a person’s life. In both ICD-10 and DSM-IV, adjustment disorders are subdivided, according to the predominant symptoms, into depressive, mixed anxiety, and depressive, with disturbance of conduct, and with mixed disturbance of emotions and conduct. DSM-IV has an additional category of ‘adjustment disorder with anxiety.’ ICD-10 has an additional category of ‘predominant disturbance of other emotion’, which includes not only adjustment disorder with anxiety, but also adjustment disorder with anger.

In ICD-10 the three types of reaction to stressful experience are classified together under ‘reactions to stress and adjustment disorders’, which is a subdivision of section F4, ‘neurotic, stress-related and somatoform disorders.’ The defining characteristics of this group of reactions to stress and of adjustment disorders are as follows:

• They arise as a direct consequence of either acute stress or continued unpleasant circumstances.

• It is judged that the disorder would not have arisen without these factors.

A different organizing principle is used in DSM-IV. Here acute stress disorder and post-traumatic stress disorder are classified as anxiety disorders, while adjustment disorders have their own place in the classification, separate from the anxiety disorders.

Additional codes in ICD-10

If any of these reactions is accompanied by an act of deliberate self-harm, another code can be added to record this fact (codes X60–X82 list 23 methods of self-harm). It is also possible to specify certain kinds of stressful event by adding a code from Chapter Z. For example, Z58 denotes problems related to employment and unemployment, and Z63 denotes problems related to family circumstances.

Coding grief reactions

In ICD-10, abnormal grief reactions are coded as adjustment disorders. Reactions to bereavement that are appropriate to the person’s culture are not included. If it is appropriate to code them as part of the description of the patient’s condition, code Z63.4 (death of a family member) can be used.

Acute stress reaction and acute stress disorder

Clinical picture

The core symptoms of an acute psychological response to stress are anxiety or depression. Anxiety is the response to threatening experiences, and depression is the response to loss. Anxiety and depression often occur together, because stressful events often combine danger and loss—an extreme example is a road accident in which a companion is killed. Other symptoms include feelings of being numb or dazed, difficulty in remembering the whole sequence of the traumatic event, insomnia, restlessness, poor concentration, and physical symptoms of autonomic arousal, especially sweating, palpitations, and tremor. Anger or histrionic behaviour may be part of the response. Occasionally there is a flight reaction—for example, when a driver runs away from the scene of a road accident.

Coping strategies and defence mechanisms are also part of the acute response to stressful events. Avoidance is the most frequent coping strategy, where the person avoids talking or thinking about the stressful events, and avoids reminders of them. The most frequent defence mechanism is denial.

Usually avoidance and denial recede as anxiety diminishes; memories of the events can be more readily accessed and the person is able to think or talk about them with less distress. This sequence allows the person to work through and come to terms with the stressful experience, although there may be continuing difficulty in recalling the details of highly stressful events.

Variations in the clinical picture

Not all responses to acute stress follow this orderly sequence, in which coping strategies and defences are maintained long enough to allow the person to function until anxiety and depression subside, and are then abandoned so that working through can occur. Not all coping strategies are adaptive—an example is the excessive use of alcohol or drugs to reduce distress. Defence mechanisms may also be of a less adaptive type, such as regression or displacement. Sometimes defence mechanisms persist for longer than is adaptive—for example, denial may persist for so long that ‘working through’ is delayed. Sometimes vivid memories of the stressful events intrude into awareness as images and flashbacks or disturbing dreams. When this state persists, the condition is called a post-traumatic stress disorder (see p. 159).

Diagnostic conventions

As noted above, ‘acute stress reaction’ in ICD-10 and ‘acute stress disorder’ in DSM-IV capture different phases of the psychological response to stress (see p. 156). The DSM-IV definition refers to cases of more clinical importance, and it is widely used. People who develop acute stress disorder are more likely to experience subsequent post-traumatic stress disorder (Brewin et al., 2003). Indeed the symptomatology of post-traumatic stress disorder is similar to that of acute stress disorder, the main difference being in the timing and duration of symptoms. However, around 50% of those who eventually develop post-traumatic stress disorder after a trauma do not meet the criteria for acute stress disorder soon after it (McNally et al., 2003).

Both systems of classification describe typical symptoms of the disorder. In DSM-IV the diagnosis of acute stress disorder requires marked symptoms of anxiety or increased arousal, re-experiencing of the event, and three of the following five ‘dissociative’ symptoms:

• a sense of numbing or detachment

• reduced awareness of the surroundings (‘being in a daze’)

• derealization

• depersonalization

• dissociative amnesia.

There must also be avoidance of stimuli that arouse recollections of the trauma, and significant distress or impaired social functioning.

In ICD-10, dissociative and other symptoms are not required to diagnose the disorder in its mild form (F43.00), but two are required for the moderate form (F43.01) and four for the severe form (F43.02), from the following list of seven:

• withdrawal from expected social interaction

• narrowing of attention

• apparent disorientation

• anger and verbal aggression

• despair and hopelessness

• inappropriate or purposeless activity

• uncontrollable and excessive grief.

The terms acute stress reaction and acute stress disorder are used only when the person was free from these symptoms immediately before the impact of the stressful event. Otherwise the response is classified as an exacerbation of pre-existing psychiatric disorder.


Rates in the population are unknown. The rate of acute stress disorder reported among survivors of motor vehicle accidents is 13% among survivors (Harvey and Bryant, 1998), and among victims of violent crime it is 19% (Brewin et al., 1999). After the Wenchuan earthquake in China, about 30% of the survivors met the DSM-IV criteria for acute stress disorder (Zhao et al., 2008).


Many kinds of event can provoke an acute response to stress—for example, involvement in a significant but brief event (e.g. a motor accident or a fire), an event that involves actual or threatened injury (e.g. a physical assault or rape), or the sudden discovery of serious illness. Some of these stressful events involve life changes to which further adjustment is required (e.g. the serious injury of a close friend involved in the same accident). Not all people who are exposed to the same stressful situation develop the same degree of response. This variation suggests that differences in constitution, previous experience, and coping styles may play a part in aetiology. However, there is little factual information available, as research has focused on the more severe and lasting post-traumatic stress disorder.


Planning for disaster

Planning is needed to ensure an immediate and appropriate response to the psychological effects of a major disaster. Such a response can be achieved by enrolling and training helpers who can support victims and are willing to be called on at short notice, and by agreeing procedures for contacting these helpers promptly. At the time of the disaster, priorities have to be decided between the needs of the victims of the disaster, those of relatives (including children), and those of members of the emergency services who may be severely affected by their experiences. The essential elements of psychological assistance for victims of disaster have been described by Alexander (2005) (see Box 8.2).


After a major incident, counselling has often taken the form known as debriefing, or Critical Incident Stress Debriefing (CISD), provided either individually or in a group. In debriefing the victim goes through the following stages after the counsellor has first explained the procedure:

Box 8.2 The principal components of psychological first aid

• Comfort and consolation

• Protection from further threat and distress

• Immediate physical care

• Helping reunion with loved ones

• Sharing the experience (but not forced)

• Linking survivors with sources of support

• Facilitating a sense of being in control

• Identifying those who need further help (triage)

• facts—the victim relates what happened

• thoughts—they describe their thoughts immediately after the incident

• feelings—they recall the emotions associated with the incident

• assessment—they take stock of their responses

• education—the counsellor offers information about stress responses and how to manage them.

Debriefing has been widely used, but current evidence suggests that single-session ‘stand-alone’ debriefing is not helpful in lowering subsequent psychological distress, and might even be harmful (for a meta-analysis, see Rose et al., 2003). The place of psychological interventions in major disasters is not without controversy. Some argue that it can isolate and pathologize the victim, and that energies should be directed towards getting individuals back to their families as soon as possible, and mobilizing normal social support structures (Summerfield, 2006).


After a traumatic event, many people talk informally to a sympathetic relative or friend, or to a member of the professional staff dealing with any physical injuries that originated during the incident. Since in most cases stress reactions will resolve with time, a policy of watchful waiting is appropriate, although it is good practice to offer a follow-up appointment after about 1 month to identify subjects whose stressful symptoms are not settling and who might therefore be at high risk of developing post-traumatic stress disorder.

More formal psychotherapy may be needed if there is no relative, friend, or professional who can assist, if the stressful circumstances cannot easily be discussed with a relative or friend (e.g. in some cases of rape), or if the response is prolonged or severe. The victim can be reassured that the condition is relatively common, and often short-lived. Advice may be needed about ways of dealing with the consequences of the traumatic event. If anxiety is severe, an anxiolytic drug may be prescribed for a day or two, and if sleep is severely disrupted a hypnotic drug may be given for one or two nights. If more formal psychotherapy is needed, there is evidence that brief trauma-focused cognitive–behaviour therapy is more effective than supportive counseling, and may help to prevent the subsequent development of post-traumatic stress disorder (McNally et al., 2003; Roberts, 2009).

Post-traumatic stress disorder

This term denotes an intense, prolonged, and sometimes delayed reaction to an intensely stressful event. The essential features of a post-traumatic stress reaction are as follows:

1. hyperarousal

2. re-experiencing of aspects of the stressful event

3. avoidance of reminders.

Examples of extreme stressors that may cause this disorder are natural disasters such as floods and earthquakes, man-made calamities such as major fires and serious transport accidents, or the circumstances of war, and rape or serious physical assault on the person. The original concept of post-traumatic stress disorder (PTSD) was of a reaction to such an extreme stressor that any person would be affected by it. Epidemiological studies have shown that not everyone who is exposed to the same extreme stressor develops PTSD; thus personal predisposition plays a part. In many disasters the victims suffer not only psychological distress but also physical injury, which may increase the likelihood of PTSD. Other predisposing factors are reviewed below in the section on aetiology.

The condition now known as PTSD has been recognized for many years, although under other names. The term PTSD originated during the study of American servicemen returning from the Vietnam War. The diagnosis meant that affected servicemen could be given medical and social help without being diagnosed as suffering from another psychiatric disorder. Similar psychological effects have been reported (under other names) among servicemen in both world wars, and among survivors of peacetime disasters such as the serious fire at the Coconut Grove nightclub in America in 1942 (Adler, 1943). For a historical review of the concept of PTSD, see Jones et al. (2003).

Other reactions to severe stress

PTSD occurs only after exceptionally stressful events, but not every response to such events is PTSD. Six months after a serious accident, major depression may actually be more frequent than PTSD (Kuhn et al., 2006). ICD-10 has a category of ‘Enduring personality changes after catastrophic experience’ (see Box 8.3). This and other conditions may occur instead of, but also as well as, PTSD. For example, of those Vietnam War veterans who met the diagnostic criteria for PTSD, 43% had at least one other diagnosis. The most frequent additional diagnoses were atypical depression, alcohol dependence, anxiety disorder, substance misuse, and somatization disorder (Mac-Farlane, 1985).

Box 8.3 ICD-10 criteria for ‘Enduring personality changes after catastrophic experience’

(A) At least two of the following:

• a permanent hostile or distrustful attitude towards the world

• social withdrawal

• a constant feeling of emptiness or hopelessness

• an enduring feeling of being on edge or being threatened without external cause

• a permanent feeling of being changed or being different from others.

(B) The change causes significant interference with personal or social functioning or significant distress.

(C) The personality change developed after the catastrophic event, and the person did not have a personality disorder prior to the event that explains the current traits.

(D) The personality change must have been present for at least 2 years, and is not related to episodes of any other mental disorder (other than PTSD) or to brain damage or disease.

Clinical picture of PTSD

The clinical features of PTSD can be divided into three groups (see Table 8.2). The first group of symptoms is related to hyperarousal, and includes persistent anxiety, irritability, insomnia, and poor concentration. The second group of symptoms centres around intrusion, and includes intense intrusive imagery of the events, sudden flashbacks, and recurrent distressing dreams. The third group of symptoms is concerned with avoidance, and includes difficulty in recalling stressful events at will, avoidance of reminders of the events, a feeling of detachment, inability to feel emotion (‘numbing’), and diminished interest in activities. The most characteristic symptoms are flashbacks, nightmares, and intrusive images, sometimes known collectively as re-experiencing symptoms.

Maladaptive coping responses may occur, including persistent aggressive behaviour, the excessive use of alcohol or drugs, and deliberate self-harm and suicide (Ahmed, 2007).

Table 8.2 The principal symptoms of post-traumatic stress disorder


Persistent anxiety



Poor concentration


Intense intrusive imagery


Recurrent distressing dreams


Difficulty in recalling stressful events at will

Avoidance of reminders of the events


Inability to feel emotion (‘numbness’)

Diminished interest in activities

Other features

Depressive symptoms are common, and guilt is often experienced by the survivors of a disaster. After some traumatic events, survivors feel forced into a painful reconsideration of their beliefs about the meaning and purpose of life. Dissociative symptoms such as depersonalization are also prominent in some patients (Lanius et al., 2010).

Onset and course

Symptoms of PTSD may begin very soon after the stressful event, or after an interval usually of days, but occasionally of months, although rarely more than 6 months. In DSM-IV, PTSD cannot be diagnosed until at least 1 month of symptomatology has elapsed; until then the condition is regarded as an acute stress disorder. However, in these circumstances it is doubtful whether the diagnosis of stress-related disorder and PTSD represents two separate conditions (Brewin et al., 2003). If the person experiences a new traumatic event, symptoms may return even if the second event is less traumatic than the original one. Many cases are persistent; about half recover within 1 year, but up to one-third do not recover even after many years (Kessler et al., 1995).


The diagnostic criteria in ICD-10 and DSM-IV are similar, although the latter assigns rather more importance to numbing. DSM-IV includes two criteria that are not present in ICD-10—symptoms must have been present for at least 1 month, and must cause significant distress or impaired social functioning. As a result of these differences, the concordance between the diagnosis of PTSD using the two sets of criteria is only 35% (Andrews et al., 1999). By convention, PTSD can be diagnosed in people who have a history of psychiatric disorder before the stressful events.

Differential diagnoses include the following:

• stress-induced exacerbations of previous anxiety or mood disorders

• acute stress disorders (distinguished by the time course)

• adjustment disorders (distinguished by the different pattern of symptoms)

• enduring personality changes after catastrophic experience.

PTSD may present as deliberate self-harm or substance abuse which have developed as maladaptive coping strategies (see above).


Estimates of PTSD in the general population have mainly been obtained from the USA. Rates in other countries are likely to differ somewhat in relation to the frequency of natural and man-made disasters in these places. In a large representative sample from the USA, Kessler et al. (1995), using DSM-III-R criteria, estimated a lifetime prevalence of PTSD of 7.8% (10.4% in women and 5.0% in men). A more recent epidemiological study conducted in the USA, employing DSM-IV diagnoses, found a rather similar lifetime risk of PTSD of 6.4% (Pietrzak et al., 2010). Estimates of the 12-month prevalence range from 1.1% in Europe (Darves-Bornoz et al., 2008) and 1.3% in Australia (Creamer et al., 2001) to 3.6% in the USA (Narrow et al., 2002)


The stressor

The necessary cause of PTSD is an exceptionally stressful event. It is not necessary that the person should have been harmed physically or threatened personally; those involved in other ways may develop the disorder—for example, the driver of a train in whose path someone has thrown himself for suicide, and the bystanders at a major accident. The authors of DSM-IV describe such events as involving actual or threatened death or serious injury or a threat to the physical integrity of the person or others. In a study of people affected by a volcanic eruption, the highest rate of PTSD was found among those who experienced the greatest exposure to the stressful events (Shore et al., 1989). Even so, not all of those most affected by the stressor developed PTSD, a finding which indicates that some form of personal vulnerability plays a part. Such vulnerability might be genetic or acquired. Epidemiological research has revealed the following findings:

• The majority of people will experience at least one traumatic event in their lifetime.

• Intentional acts of interpersonal violence, in particular combat and sexual assault, are more likely to lead to PTSD than accidents or disasters.

• Men tend to experience more traumatic events in general than women, but women experience more events that are likely to lead to PTSD (e.g. childhood sexual abuse, rape, and domestic violence).

• Women are also more likely to develop PTSD in response to a traumatic event than men. This enhanced risk is not explained by differences in the type of traumatic event.

Genetic factors

Studies of twins suggest that differences in susceptibility are in part genetic. True et al. (1993) studied 2224 monozygotic and 1818 dizygotic male twin pairs who had served in the US armed forces during the Vietnam War. After allowance had been made for the amount of exposure to combat, genetic variation accounted for about one-third of the variance in susceptibility to self-reported PTSD. Self-reported childhood and adolescent environment did not contribute substantially to this variance. The risk of PTSD is increased by a family history of psychiatric disorder, which may reflect genetic factors. Attempts to identify specific genes important in aetiology have not led to consistent findings (Cornellis et al., 2010).

Other predisposing factors

The individual factors that increase vulnerability to the development of PTSD have been summarized by Ahmed (2007). They include the following:

• personal history of mood and anxiety disorder

• previous history of trauma

• female gender

• neuroticism

• lower intelligence

• lack of social support.

Neurobiological factors

Research to date on the neurobiology of PTSD has focused on monoamine neurotransmitters and the hypothalamic–pituitary–adrenal (HPA) axis, both of which are involved in mediating defensive responses to stressful events. In addition, brain imaging studies have implicated changes in the hippocampus, a brain region that is important in memory formation, and the amygdala, which plays a role in non-conscious emotional processing (see Box 8.4). These findings suggest that hippocampal dysfunction prevents adequate memory processing, while increased activity in noradrenergic innervation of the amygdala increases arousal and facilitates the automatic encoding and recall of traumatic memories. Functional imaging studies in PTSD suggest overactivity of the amygdala in the context of decreased regulatory control of limbic regions by the ventromedial prefrontal cortex (Krystal and Neumeister, 2009). However, some have argued that forms of PTSD characterized by prominent dissociative symptoms (emotional detachment, derealization, and depersonalization) may in fact be characterized by excessive corticolimbic control (Lanius et al., 2010).

Psychological factors

Fear conditioning

Some patients with PTSD experience vivid memories of the traumatic events in response to sensory cues such as smells and sounds related to the stressful situation. This finding suggests that classical conditioning may be involved, as well as failure to extinguish conditioned responses (Jovanovic and Ressler, 2010).

Box 8.4 Neurobiological abnormalities in PTSD

Hypothalamic–pituitary–adrenal (HPA) axis

Evidence, albeit somewhat contradictory, for low plasma cortisol levels and increased sensitivity to dexamethasone suppression. Increased levels of corticotropin-releasing hormone in CSF. Longitudinal sampling suggests dysregulation of the HPA axis with a general increase in lability following environmental stress (Pervanidou and Chrousos, 2010). Higher levels of glucocorticoid receptors may predict the development of PTSD in response to trauma (van Zuiden et al., 2011).


Increased sympathetic tone. Increased startle response. Increased levels of 3-methoxy-4-hydroxy-phenylglycol (MHPG) in CSF. Increased anxiety response to noradrenaline challenge. Decreased levels of neuropeptide Y at baseline and in response to noradrenaline challenge (Krystal and Neumeister, 2009).

Brain imaging

Smaller volume of the hippocampus (which may be a vulnerability factor), overactivity of the amygdala in response to traumatic psychological stimuli (Kitayama et al., 2005; Krystal and Neumeister, 2009).

Cognitive theories

These suggest that PTSD arises when the normal processing of emotionally charged information is overwhelmed, so that memories persist in an unprocessed form in which they can intrude into conscious awareness. In support of this idea, patients with PTSD tend to have incomplete and disorganized recall of the traumatic events. Individual differences in response to the same traumatic events are explained as being due to differences in the appraisal of the trauma and of its effects. Similarly, differences in the appraisal of the early symptoms may explain why these symptoms persist for longer in some individuals. Negative interpretations of intrusive thoughts (e.g. ‘I am going mad’) after road accidents predict the continuing presence of PTSD after 1 year. The cognitive model of PTSD has been reviewed by Ehlers and Clark (2008).

Psychodynamic theories

These emphasize the role of emotional development in determining individual variations in the response to severely stressful events. The general approach is plausible, and is supported by the fact that factors such as positive self-esteem, trust, and secure attachment increase resilience and decrease the risk of experiencing PTSD following trauma (see Ahmed, 2007).

Maintaining factors

As noted above, symptoms of PTSD may be maintained in part by negative appraisals of the early symptoms. Other suggested maintaining factors include avoidance of reminders of the traumatic situation (which prevents deconditioning and cognitive reappraisal), suppression of intrusive memories (which is known to make them more likely to recur), and rumination (Murray et al., 2002).


This should include enquiries about the nature and duration of symptoms, previous personality, and psychiatric history. When the traumatic events have included head injury (e.g. in an assault or transport accident), a neurological examination should be performed. Feelings of anger and thoughts of self-harm are common in PTSD, and an appropriate risk assessment needs to be carried out. Secondary complications such as substance misuse may require treatment in their own right (see Chapter 17).



The treatment of PTSD can be difficult, particularly once symptoms have been established for more than a year. The general approach to more short-lived cases is to provide emotional support, to encourage recall of the traumatic events to integrate them into the patient’s experience, and to facilitate working through the associated emotions. Treatment may also need to deal with the person’s feelings of guilt about perceived shortcomings in responding during the events, grief, and guilt about surviving when others have died. There may be existential concerns about the meaning and purpose of life and death. Victims of personal assault or rape have additional concerns (see p. 164).

Cognitive–behavioural treatment

Where symptoms are severe or long-standing, cognitive–behaviour therapy is the most appropriate treatment. This treatment has several components:

• information about the normal response to severe stress, and the importance of confronting situations and memories related to the traumatic events

• self-monitoring of symptoms

• exposure to situations that are being avoided

• recall of images of the traumatic events, to integrate these with the rest of the patient’s experience. When first recalled these images are often fragmentary and are not clearly related in time to the other contents of memory

• cognitive restructuring through the discussion of evidence for and against the appraisals and assumptions

• anger management for people who still feel angry about the traumatic events and their causes.

A meta-analysis of psychotherapy studies of PTSD suggests that cognitive–behavioural treatments have a therapeutic effect size of 1.65 when compared with an inactive control, and 1.01 when compared with relaxation and supportive therapies. An effect size of 1 corresponds to an improvement of one standard deviation on the relevant symptomatic measure, and effect sizes of > 1 indicate a large treatment effect (see p. 119). At the end of psychological treatment, around 55% of patients no longer meet the criteria for PTSD, although many are still symptomatic. Although these results are encouraging, further work is required to show that similar benefit can be obtained in everyday clinical settings (Bradley et al., 2005).

Eye movement desensitization and reprocessing was designed for the treatment of PTSD (see p. 583). Treatment trials using this technique in subjects with PTSD have shown similar effect sizes to those obtained with cognitive–behaviour therapy, but the evidence base is not as large (Bradley et al., 2005). Some have questioned whether the eye movements associated with this treatment add any specific therapeutic value to the element of exposure.

There is little evidence for the effectiveness of other psychotherapies, such as psychodynamic therapy and hypnotherapy. For a review, see the National Institute for Health and Clinical Excellence (2005b).


Anxiolytic drugs such as benzodiazepines should be avoided in patients with established PTSD, because prolonged use may lead to dependence. A number of antidepressant drugs have shown efficacy in clinical trials, including selective serotonin reuptake inhibitors (SSRIs), serotonin and noradrenaline reuptake inhibitors (SNRIs), tricyclic antidepressants (TCAs), and monoamine oxidase inhibitors (MAOIs). There are also more preliminary data supporting the efficacy of mirtazapine, and augmentation with atypical antipsychotic drugs such as olanzapine may have a place in treatment-resistant patients (Ravindran and Stein, 2009). However, a meta-analysis conducted by the National Institute for Health and Clinical Excellence (2005b) indicated that in patients with PTSD, drug treatment has a lower effect size than structured psychotherapy, and medication should not therefore be a first-line treatment unless the patient expresses a preference for it, or psychotherapy is not available or is ineffective. However, where patients with PTSD have a significant comorbid depressive disorder, antidepressant treatment should be considered as an adjunct to psychotherapy. For reviews of the treatment of PTSD, see the National Institute for Health and Clinical Excellence (2005b) (see also Box 8.5), Bradley et al. (2005), and Ravindran and Stein (2009).

Box 8.5 National Institute for Health and Clinical Excellence (NICE) guidance for the management of PTSD and related traumatic syndromes

1. The routine use of a brief screening instrument for PTSD at 1 month post incident should be considered for all people who have been involved in a major disaster.

2. Where symptoms are mild and have been present for less than 4 weeks after the trauma, watchful waiting—as a way of managing the difficulties presented by individual sufferers—should be routinely considered by healthcare professionals. A follow-up contact should be arranged within 1 month.

3. Trauma-focused cognitive–behavioural therapy should be considered for those with severe post-traumatic symptoms during the first month after the traumatic event. All PTSD sufferers should be offered a course of trauma-focused psychological treatment. These treatments are usually given individually on an outpatient basis.

4. Drug treatments for PTSD should not be used as a routine first-line treatment in preference to a trauma-focused psychological therapy.

5. Drug treatments (paroxetine, mirtazapine, amitriptyline, phenelzine) should be considered for the treatment of PTSD where a sufferer expresses a preference not to engage in a trauma-focused psychological treatment.

Adapted from the National Institute for Health and Clinical Excellence (2005b).

Response to special kinds of severe stress

Rape and physical assault

Victims of rape or physical assault experience acute reactions to stress, PTSD, anxiety and depressive disorders, and psychosexual dysfunction. PTSD is the most frequent of these consequences. In one study of women victims of rape, 94% had acute stress disorder after the assault, and 47% had the symptoms 3 months later (Rothbaum et al., 1992).

As well as experiencing symptoms of PTSD, victims of rape and assault feel humiliated, ashamed, and vulnerable to further attack. They lose confidence and self-esteem, question why they were chosen as victims, and blame themselves for putting themselves in unnecessary danger. To these problems are added issues of betrayal and secrecy when the rapist is a family member or a friend. The victim may experience problems in trusting, persistent anger and irritability, and excessive dependence. These problems were first described among women victims of rape, but similar difficulties have been described among male victims of sexual assault (Peterson et al., 2011).

Problems are more likely to persist when there has been an actual or perceived threat to life, previous psychological and social problems, past victimization, particularly abuse in childhood, past psychiatric illness or substance abuse, or a lack of social support (Adshead and Ferris, 2007).


Social support is important in providing opportunities for the victim to talk over the problem and to regain self-esteem. Specific treatment is similar to that for other kinds of PTSD, including prolonged exposure by reliving the events in imagination, with additional emphasis on overcoming feelings of vulnerability and self-blame (Vickerman and Margolin, 2009).

War and other armed conflict

‘Shell shock’, ‘battle fatigue’, and ‘war neurosis’ were terms used during the First World War to describe psychological reactions to battle among British and American servicemen. Most of the reactions appear to resemble cases now diagnosed as PTSD; others seem to have resembled panic disorder or depressive disorders. Cases with panic attacks and concerns about the heart, now diagnosed as panic disorder (see p. 195), were known then as Da Costa’s syndrome or disorderly action of the heart. Army psychiatrists were few in number, and were unable to deal with the many cases. Moreover, their experience of mental hospital work with severely ill patients did not equip them to treat these reactions to battle. Therefore patients with shell shock were treated mainly by neurologists or psychologists. W. H. Rivers, William Brown, and William McDougall were British psychologists who treated shell shock during the First World War, and used this experience to write influential books on medical psychology in the years after the war (Rivers, 1920; McDougall, 1926; Brown, 1934).


At first, shell shock was treated with the methods in use at the time for neurasthenia (see p. 91), namely rest, isolation, massage, and diet, but these methods had a low success rate. Hypnosis achieved some dramatic cures, but was not generally effective. Medical psychologists tried psychotherapeutic methods advocated by Freud, including the recall of stressful events to remove repression and the expression of associated emotion. There was an increasing emphasis on early treatment, keeping the soldier with his colleagues at the front with the explicit aim of rapid return to action (‘the soldier’s interest and the army’s interest are the same’). Psychotherapy had to be combined with military drill to maintain general fitness and morale. This combined treatment led to improved results. Chronicity increased dramatically if soldiers were evacuated from the theatre of war.

These general principles of early treatment, abreaction, and maintenance of fitness and morale were adopted in the Second World War and in subsequent conflicts. Abreaction with anxiolytic drugs was used widely in the Second World War (Sargant and Slater, 1940). In subsequent conflicts it has been reported that, with immediate counselling (without drug-induced abreaction), about 70% of ‘battle-shock’ personnel can be returned to their units within 5 days (Brandon, 1991). The treatment of shell shock in the British army in the First World War has been described by Stone (1985).

Problems of refugees and victims of torture

Refugees may have experienced a wide range of traumatic events, including the following:

• the conditions of war

• loss by death or separation of relatives and friends

• loss of home and possessions

• physical injury (including brain injury) either from the actions of war or from assault, rape, or torture, and the witnessing of violence to others.

Those involved may develop any of the reactions to stressful events, especially PTSD and depressive disorders. These conditions have been identified in refugees from many cultures, although the presenting complaints may differ somewhat in people from different cultures, with more emphasis on physical than on psychological symptoms among people from non-Western countries. Victims of torture often experience PTSD as well as the physical consequences of the experience. However, it is important to remember that stressors related to a refugee’s current situation can be just as difficult to deal with as those that led to flight in the first place (see Box 8.6). In a systematic review, Fazel et al. (2005) estimated that of refugees re-settled in Western countries, about 10% met the diagnostic criteria for PTSD, 5% had major depression, and 4% had generalized anxiety disorder. In many individuals these disorders overlapped.


Treatment should combine physical and psychiatric care. The latter should be introduced carefully, since it may be resisted as stigmatizing, not only by the refugee but also by aid workers. Special care is needed to establish a trusting relationship. Practical help is often an essential preliminary to engagement. Many refugees have problems related to separation, bereavement, and loss of material possessions, so it is important not to focus narrowly on PTSD. Victims of torture may need additional help to cope with feelings of personal humiliation and of remorse for the suffering of others. Health beliefs and understanding of the ‘normal’ psychological response to stress may well differ between a health worker from one culture and a refugee from another. Therefore it is necessary to consider for each individual how far formal psychological therapies may be indicated and whether or not they should be presented in the language of mental illness and trauma (Tribe, 2002; McColl et al., 2008).

Box 8.6 Some stressful issues faced by refugees



Human rights abuses

Persecution on grounds of politics, religion, gender, or ethnicity

Resultant losses








Plans for the future

Issues in country of asylum

Psychological and practical adjustment

Uncertain future

Traumatic life events

Social exclusion and poverty


Stereotyping by host country

Unknown cultural traditions

From Tribe (2002).

Care is needed when working through interpreters. If possible, they should not be family members, community elders, or others to whom the refugee is unlikely to speak of shameful experiences. This point is especially important in situations in which women may have experienced sexual assaults, for these may bring shame on the whole family. Ideally, such problems should be dealt with by a female mental health professional who understands the patient’s language and culture, but this may be difficult to arrange.

For further information about the psychiatric problems of refugees, see Lustig et al. (2004) and McColl et al. (2008).

Adjustment disorders

This term refers to the psychological reactions that arise in relation to adapting to new circumstances. Such circumstances include divorce and separation, a major change of work and abode (e.g. transition from school to university, or migration), and the birth of a handicapped child. Bereavement, the onset of a terminal illness, and sexual abuse are associated with special kinds of adjustment, which are discussed below.

Clinical features

The symptoms of an adjustment disorder include anxiety, worry, poor concentration, depression, and irritability, together with physical symptoms caused by autonomic arousal, such as palpitations and tremor. There may be outbursts of dramatic or aggressive behaviour, single or repeated episodes of deliberate self-harm, or the misuse of alcohol or drugs. The onset is more gradual than that of an acute reaction to stress, and the course is more prolonged. Social or occupational function is impaired. The impairment in social or occupational function as well as the intensity of distress is what distinguishes adjustment disorder from normal adaptive reactions.

Stressful life events may precipitate depression, anxiety, schizophrenia, and other psychiatric disorders. For this reason, the diagnosis of adjustment disorder is not made when diagnostic criteria for another psychiatric disorder are met. In practice, therefore, the diagnosis is usually made by excluding an anxiety or depressive disorder. A further requirement for diagnosis is that the disorder starts soon after the change of circumstances. Both ICD-10 and DSM-IV require that the disorder starts within 3 months, and ICD-10 indicates that it usually starts within 1 month. An essential point is that the reaction is understandably related to, and in proportion to, the stressful experience when account is taken of the patient’s previous experiences and personality. Once the stressor or its consequences are removed, the symptoms resolve within 6 months.

Diagnostic conventions

As explained on p. 156, in ICD-10 adjustment disorders are divided into depressive reactions, mixed anxiety and depressive reactions, reactions with disturbance of other emotions, and reactions with disturbed conduct with or without emotional disturbance. DSM-IV lists six types of adjustment disorder (see Table 8.2).


The prevalence of adjustment disorder in the community has been little studied. A community study in older people found that the prevalence of adjustment disorders was about 2% (Maercker et al., 2008). It is presumed that in certain settings, such as the general hospital and primary care, prevalence rates are increased. For example, in studies of general hospital inpatients referred for psychiatric consultation, rates of adjustment disorder ranged from 11.5% to 21.5% (Strain et al., 2009). High rates are also seen in people who make suicide attempts (see p. 434).


Stressful circumstances are the necessary cause of an adjustment disorder, but individual vulnerability is also important, because not all people who are exposed to the same stressful circumstances develop an adjustment disorder. The nature of this vulnerability is unknown; it seems to vary from one person to another, and may relate in part to previous life experiences.


Clinical experience suggests that most adjustment disorders last for several months, and a few persist for years. There is little systematic follow-up information, although Andreasen and Hoenck (1982) reported that while the prognosis is good for adults, the majority of adolescents with adjustment disorder develop psychiatric disorders in adult life.


Treatment is designed to help the patient to resolve the stressful problems if this is possible, and to aid the natural processes of adjustment. The latter is done by reducing denial and avoidance of the stressful events, encouraging problem solving, and discouraging maladaptive coping responses. Anxiety can usually be reduced by encouraging the patient to talk about the problems and to express their feelings. Occasionally, an anxiolytic or hypnotic drug is needed for a few days.

Problem-solving counselling (see p. 576) encourages the patient to seek solutions to stressful problems, and to consider the advantages and disadvantages of various kinds of action. The patient is then helped to select and implement a course of action to solve the problem. If this action succeeds, another problem is considered. If the first attempt fails, another approach to the original problem is tried. If problems cannot be resolved, the patient is encouraged to come to terms with them. Maina et al. (2005) reported that in patients with adjustment disorder, both brief dynamic and supportive psychotherapy were more effective than a waiting-list control. For a review of adjustment disorder, see Casey (2009).

Special kinds of adjustment

Adjustment to physical illness and handicap

Appraisal of illness

Adjustment to illness cannot be understood simply in terms of the facts about the disease and its objective consequences. Adjustment depends on the patient’s beliefs about their disorder and its effects on their life—in other words, on their appraisal of their illness. This appraisal may be similar to that of the professionals who are treating them, or it may be very different because it is based on false information or on emotions rather than on facts, or influenced by cultural beliefs. The appraisal may be reinforced by members of the family who share the patient’s views, or it may be contradicted by them, thus adding to the patient’s distress. Two terms are much used in the discussions of adjustment to illness and handicap—illness behaviour and the sick role. These terms will be considered next.

Illness behaviour

Mechanic (1978) suggested the term illness behaviour to describe behaviour associated with adjustment to physical or mental disorder, whether adaptive or not. Illness behaviour includes consulting doctors, taking medicines, seeking help from relatives and friends, and giving up inappropriate activities. These behaviours are adaptive in the early stages of illness, but may become maladaptive if they persist into the stage of convalescence when the patient should be becoming independent. Illness behaviour results from the person’s conviction that he is ill rather than from the objective presence of disease, and it may develop when no disease is present. Illness behaviour without disease is an important problem in general practice, and once firmly established it is difficult to treat. The concept of illness behaviour overlaps with that of the sick role (described below), but the two are described separately because they have different origins.

The sick role

Society bestows a special role on people who are ill. Parsons (1951) called this the sick role, which is made up of two privileges and two duties:

• exemption from certain social responsibilities

• the right to expect help and care from others

• the obligation to seek and cooperate with treatment

• the expectation of a desire to recover.

While the person is ill, the sick role is adaptive. If they continue in the sick role after the illness is over, recovery is delayed as they continue to avoid responsibilities and depend on others instead of becoming independent.

Adjustment to the onset of physical illness

When a person becomes physically ill, they may feel anxious, depressed, or angry. Usually this emotional reaction is transient, subsiding as the patient comes to terms with the new situation. As in other adjustment reactions, denial or minimization can protect the patient against overwhelming anxiety when the diagnosis is first known. Although helpful in this way, denial can be maladaptive—in the early stage of illness it may lead to delay in seeking help, and at a later stage it may lead to poor compliance with treatment. Other coping strategies can be divided into emotion-reducing and problem-solving groups (see p. 154). Coping strategies that reduce emotion are often appropriate in the early stages of illness, but should give way to problem-solving coping. Coping may fail when demands are very great, or when coping resources are limited either in the long term, or as a temporary result of disease of, or trauma to, the brain.

Physical illness as a direct cause of psychiatric symptoms

As well as acting as a stressor, physical illness may induce psychiatric symptoms directly. Anxiety, depression, fatigue, weakness, weight loss, or abnormal behaviour may all be caused directly by physical disorders; common examples are listed in Chapter 15 (Table 15.5). Similarly, sexual function may be impaired by physical illness or its treatment (see p. 367). Any of these symptoms may be the reason for referral, and psychiatrists should always be alert for the possibility of undetected physical illness in their patients.

Psychiatric symptoms due to treatments for physical illness

Some drugs that are used in the treatment of physical illness may affect mood, behaviour, and consciousness. The drugs most likely to have these effects are listed in Chapter 15 (Table 15.6).

Help for people who are adjusting to physical illness

Most people adjust well to physical illness, but when adjustment is slow and incomplete, psychological treatment may be needed. This treatment need not be complicated, and can usually be provided effectively by the general practitioner or the hospital doctors or nurses who are managing the physical illness. Generally, the psychiatrist has a role in treating only the most severe problems, or in supporting the medical and nursing staff.

The first step is to identify patients who are adjusting badly (i.e. failing to cope). This is generally done by the professional staff who are managing the physical illness. They can do this most easily by looking out for patients who are progressing less well than would be expected on the basis of the severity of the disease. Mood disorders are a common cause of slow progress, but some are dismissed as normal responses to the problems of the illness. Screening questionnaires can be used to detect mood disorders among this group, but the results should be checked at least by a brief interview, which should be conducted if possible in surroundings in which the patient’s replies will not be overheard. Generally, one or more members of the family should be interviewed to obtain information about the patient’s previous adjustment to problems and illness, and to discover how the family views the illness.

Some patients require medication, but for many counselling (see p. 575) is more appropriate. Counselling requires a trusting relationship with the patient, and this in turn requires adequate time for the interviews. Counselling begins with an explanation of the nature of the illness and its treatment. The patient is then helped to accept the implication of the diagnosis, to adjust to the illness, and to give up any maladaptive behaviours, such as excessive dependence on others or denial of the need for treatment. Graded activities, motivational interviewing, and anger management may be useful in some cases. More structured psychological treatment in the form of cognitive–behaviour therapy may also be useful (Halford and Brown, 2009).

If the reaction to physical illness is an anxiety or depressive condition, treatment appropriate to that disorder should be given (see Chapters 9 and 10).

Adjustment to terminal illness

Among patients dying in hospital, about 50% have emotional symptoms of anxiety, depression, anger, or guilt (Meyer et al., 2003). Determinants of emotional reactions include the patient’s personality, and the amount and quality of support from family, friends, and carers. Understandably, emotional reactions are more common among young dying patients than among the elderly. Surveys indicate that loss of independence and dignity and the management of pain are major concerns (Clarke and Seymour, 2010).


Anxiety may be provoked by the prospect of severe pain, disfigurement, or incontinence, by fear of death, and by concerns about the future of the patient’s family. Families and carers sometimes try to spare the patient anxiety by concealing the truth about the condition. Since most patients become aware of the diagnosis, attempts at concealment only serve to increase their fear of possible consequences of the disease, such as pain or incontinence.


Depression may be provoked by the prospect of separation from family and friends and the loss of valued activities. Changes in physical appearance caused by the illness, the effects of surgery, and the debilitating effects of radiotherapy are other causes of low mood.

Guilt and anger

Some patients experience guilt because they believe that they are making excessive demands on relatives or friends. Patients with religious beliefs may believe that illness is a punishment for previous wrongdoing. Anger may be felt about the unjustness of impending death; this anger may be displaced on to doctors, nurses, and relatives, making care more difficult (see below).

Defence mechanisms

The defence mechanisms that are most often observed in dying patients are denial, dependency, and displacement.

• Denial is usually the first reaction to the news of fatal illness. It may be experienced as a feeling of disbelief, and may lead to an initial period of calm. Denial diminishes as the patient becomes reconciled to the illness. It may return as the disease progresses, and the patient may again behave as if they are unaware of the nature of the illness.

• Regression in the form of dependency is adaptive in the early stages of severe physical illness, when the patient needs to comply passively with treatment. Excessive or prolonged dependency makes subsequent treatment more difficult, and increases the burden on the family. A further stage of dependency may be appropriate as the patient nears death.

• Displacement is often of anger, which may be directed towards staff and relatives, who may not understand this reaction, so find it difficult to tolerate. As a result they may spend less time with the patient, thereby increasing his or her feelings of despair.

Denial, dependency, and displacement are usually followed by acceptance. The doctor’s aim should be to help the patient to reach this acceptance before the final stage of the illness. This is more likely to be achieved if there is good communication between the patient, the staff caring for them, and the relatives.

Psychological symptoms

Psychological symptoms induced by the disease or by its treatment may add to the patient’s distress. The more frequent associations between disease and psychological symptoms are summarized in Table 15.5 (p. 384). There is a particularly strong association between dyspnoea and anxiety. The associations between drug treatment and psychological symptoms are summarized in Table 15.6 (p. 384).


Usually dying patients are helped to adjust by the staff who are managing the physical illness. Psychiatrists are called upon only when there are special problems (see below), or to assist with staff support and training.

The aims of treatment

According to Hackett and Weissman (1962), the aim of treating the dying patient should be to achieve an ‘appropriate death.’ By this they meant that the patient should be relatively free from pain, should operate on as effective a level as possible, should recognize and resolve any remaining conflicts, should satisfy as far as possible their remaining wishes, and should be able to yield control to others in whom they have confidence.

Kubler-Ross (1969) formulated the aims in different terms, and described five phases of psychological adjustment to death. The phases do not necessarily occur in the same sequence, and some of them may not occur at all, but they are a useful guide for professionals who are helping dying patients. They can be summarized as follows:

• denial and isolation

• anger

• partial acceptance (‘bargaining for time’)

• depression

• acceptance.

Reducing symptoms

Adequate control of pain and breathlessness and the reduction of confusion due to delirium are particularly important. Anxiety and depression may diminish as pain and breathlessness are controlled. The causes of delirium are listed on p. 318. Among dying patients, important remediable causes are dehydration, the side-effects of drugs, secondary infection, cardiac or respiratory failure, and hypercalcaemia.

Helping the patient to adjust

It is essential to establish a good relationship with the patient so that they can talk about their problems and ask questions. The nature of the illness should be explained honestly and in simple language. Sometimes doctors are apprehensive that such an explanation will increase the patient’s distress. Although excessive detail given unsympathetically can have this effect, it is seldom difficult to decide how much to say about diagnosis and prognosis, provided that patients are allowed to lead the discussion, express their worries, and say what they want to know. If patients ask about the prognosis, they should be told the truth; evasive answers undermine trust in the carers. Hinton’s pivotal study (Hinton, 1971), which was undertaken at a time when most doctors thought that patients should not be told the truth about their prognosis, found that they would themselves insist on being told. If a patient does not seem to wish to know the full extent of their problems, it is usually better to save this information until later. At an appropriate stage the patient should be told what can be done to make their remaining time as comfortable as possible. While the whole account should be truthful, the amount that is disclosed on a single occasion should be judged by the patient’s reactions and by their questions. If necessary, the doctor should be prepared to return for further discussion when the patient is ready to continue. It is important to bear in mind that most dying patients become aware of their prognosis whether or not they are told directly, because they infer the truth from the behaviour of those who are caring for them. They notice when answers to questions are evasive and when people avoid talking to them. Patients who are anxious, angry, or despairing need to be able to express these feelings and to discuss the ideas that induce them.

Informing the staff

All of the staff should be made aware of the information that has been given to the patient, otherwise conflicting advice and opinions may be offered. If all those involved know what has been said, they will feel more at ease when talking to the patient. Otherwise they will draw back from the patient, isolating them, and thereby increasing their difficulty in adjusting.

Informing and supporting the relatives

Relatives need to know what has been said to the patient so that they will feel less ill at ease when talking to them. They may need as much help as the patient. They may become anxious and depressed, and they may respond with guilt, anger, or denial. Such reactions make it difficult for them to communicate helpfully with the patient or the staff. Relatives need information, and opportunities to talk about their feelings and to prepare for the impending bereavement, otherwise the patient and their family may become increasingly distant and alienated.

Special services

In many hospitals, specialist nurses work with the family doctor and with the hospital staff who are caring for dying patients. These nurses are trained in the psychological as well as physical care of the dying. Sometimes care is provided in hospices, where it is possible to provide close attention to the details of care that improve quality of life for the dying person. These hospices care for patients when home care is impractical, and provide periods of respite care to relieve those who are caring for the patient at home.

Referral to a psychiatrist

Referral to a psychiatrist is appropriate when psychiatric symptoms or behaviour disturbance are severe. The referrals are concerned with the assessment and management of:

• depressed patients, to determine the cause and whether the patient requires medication or more structured psychotherapy

• uncommunicative patients who will not talk about the illness

• uncooperative patients who do not accept the social restrictions imposed by the illness, will not make appropriate plans, or will not take the necessary decisions

• long-standing problems that are made worse by the illness and that are related to personality or family conflicts

• other symptoms: although anxiety and delirium are common, these problems are more often dealt with appropriately by medical staff than referred to a psychiatrist; the exception is delirium with paranoid symptoms.

Management of depressive disorders

Depressive disorders may be caused by symptoms such as pain or breathlessness, and all such symptoms should be treated appropriately. Any drugs that can cause depression (see Table 15.6 on p. 384) should be reviewed and, if possible, given at a lower dose or replaced. Some symptoms of depressive disorder are difficult to evaluate in patients with advanced cancer, as weight loss, anorexia, insomnia, loss of interest, and fatigue may be caused by the physical illness. Early-morning wakening, extreme hopelessness, and self-blame are more reliable guides to diagnosis. Suicidal ideation should be assessed carefully. If counselling and improved medical management do not improve the low mood, antidepressant drugs should be prescribed with careful supervision. The starting dose should be small, and medication should be changed if necessary to find a compound that is well tolerated (Dein, 2003).

Liaison with medical and nursing staff

Liaison with medical and nursing staff is important. Often these staff can provide treatment when the psychiatrist has formulated a plan. For further information about the care of the dying, see Billings (2000).

Grief and adjustment to bereavement


Although the words ‘bereavement’, ‘mourning’, and ‘grief’ are sometimes used interchangeably, they have separate meanings which incorporate distinctions that are useful in psychiatry.

• Bereavement is the loss through death of a loved person.

• Grief is the involuntary emotional and behavioural response to bereavement.

• Mourning is the voluntary expression of behaviours and rituals that are socially sanctioned responses to bereavement. These behaviours and rituals differ between societies and between religious groups both in their form and in their duration.

The systems of classification do not make these distinctions in consistent ways. In ICD-10, bereavement is coded appropriately as Z63.4—that is, as one of the ‘factors influencing health status and contact with health services.’ In DSM-IV, however, bereavement is coded as a ‘condition that may be the focus of clinical attention’—thus the term is used to denote the response to bereavement, rather than the event itself. ICD-10 codes grief under adjustment disorders, but uses the term ‘grief reaction.’ Mourning, which is a form of social behaviour, is not a disorder and, appropriately, is not listed in the index to either classification. In this chapter the term bereavement reaction is used to denote all responses to bereavement, both normal and abnormal. Normal reactions are referred to as grief, while abnormal reactions include abnormal (or pathological) grief, and depressive disorders.


Grief is a continuous process, but for clarity can be described as having three stages (see Table 8.3).

The first stage lasts from a few hours to several days. There is denial, which is manifested as a lack of emotional response (‘numbness’), often with a feeling of unreality, and incomplete acceptance that the death has taken place. The bereaved person may be restless, as if searching for the dead person.

The second stage usually lasts from a few weeks to about 6 months, but may be much longer. There may be extreme sadness, weeping, loneliness, and often overwhelming waves of yearning for the dead person. Anxiety is common; the bereaved person is anxious and restless, sleeps poorly, lacks appetite, and may experience panic attacks. Many bereaved people feel guilt that they failed to do enough for the deceased. Some feel anger and project their feelings of guilt, blaming doctors or others for failing to provide optimal care for the person who has died. Many bereaved individuals have a vivid experience of being in the presence of the dead person, and about one in ten experience brief hallucinations. The bereaved individual is preoccupied with memories of the dead person, sometimes in the form of intrusive images. Withdrawal from social relationships is frequent, and complaints of physical symptoms are common.

Table 8.3 Normal grief reaction

Stage 1: hours to days

Denial, disbelief


Stage 2: weeks to 6 months

Sadness, weeping, waves of grief

Somatic symptoms of anxiety


Poor sleep

Diminished appetite

Guilt, blame of others

Experience of a presence

Illusions, vivid imagery

Hallucinations of the dead person’s voice

Preoccupation with memories of the deceased

Social withdrawal

Stage 3: weeks to months

Symptoms resolve

Social activities resumed

Memories of good times

Symptoms may recur at anniversaries.

In the third stage, these symptoms subside and everyday activities are resumed. The bereaved person gradually comes to terms with the loss and recalls the good times shared with the deceased in the past. Often there is a temporary return of symptoms on the anniversary of the death.

Although these stages are a useful guideline, individual responses vary, and no one feature is universal.

Abnormal or pathological grief

Grief is considered to be abnormal if it is unusually intense, unusually prolonged, delayed, or inhibited or distorted. The criterion for abnormal intensity is that the symptoms meet the criteria for a depressive disorder. The criterion for abnormal duration is that the response lasts for more than 6 months. The usual criterion for delay is that the first stage of grief has not occurred by 2 weeks after the death. In all of these forms of grief, persistent avoidance of situations and of other reminders of death are common.

Abnormally intense grief. Depressive symptoms are a frequent component of normal grief, and up to 35% of bereaved people meet the criteria for a depressive disorder at some time during their grieving (Zisook and Shear, 2009). Most of these depressive disorders resolve within 6 months, but about 20% persist for longer. It might be argued that if about one-third of bereaved people meet the criteria for depressive disorder at some time, the threshold has been set too low. However, people who meet the criteria for a depressive disorder are more likely to have poor social adjustment, to visit doctors frequently, and to use alcohol (Zisook and Shear, 2009). Therefore it is of practical value to use the criterion and to record the additional diagnosis of a depressive disorder in these cases. If there is doubt whether depressive disorder should be recorded, particular attention should be paid to symptoms of retardation and global loss of self-esteem (clearly greater than regret about omissions of care during the terminal illness), because these features are seldom present in uncomplicated grief. It is also important to assess whether suicidal feelings are present.

Prolonged grief. As explained above, prolonged (or chronic) grief is often defined as grief that lasts for more than 6 months. Instead of the normal progression, symptoms of the first and second stages persist. However, it is difficult to set a precise limit to normal grief, and complete resolution may take much longer. One study found that only a minority of widows had ceased to grieve a year after the death (Parkes and Brown, 1971). Prolonged grief may be associated with a depressive disorder, but can occur without it. Prolonged grief has also been described as ‘traumatic grief’, in which subjects show persistent searching, yearning, disbelief regarding the death, and preoccupation with thoughts of the deceased (Boelen et al., 2003).

Delayed grief. By convention, delayed grief is said to occur when the first stage of grief does not appear until more than 2 weeks after the death. It is said to be more frequent after sudden, traumatic, or unexpected deaths.

Inhibited and distorted grief. The term inhibited grief refers to a reaction that lacks some normal features. Distorted grief refers to features (other than depressive symptoms) that are either unusual in degree (e.g. marked hostility, overactivity, and extreme social withdrawal), or else unusual in kind (e.g. physical symptoms that were part of the last illness of the deceased). These distorted presentations were described by Lindemann (1944) in a study of survivors of a fire in a nightclub. In all these forms of grief, persistent avoidance of situations and of other reminders of death is common.

Causes of abnormal grief

Abnormal grief is generally thought to be more likely to occur when:

• the death was sudden and unexpected

• the bereaved person had a very close, or dependent, or ambivalent relationship with the deceased

• the survivor is insecure, or has difficulty in expressing their feelings, or has suffered a previous psychiatric disorder

• the survivor has to care for dependent children and so cannot show their grief easily.

Morbidity after bereavement

Several studies (reviewed by Stroebe et al., 2007) have shown an increased rate of mortality among bereaved spouses and other close relatives, with the greatest increase being in the first 6 months after bereavement. Most studies report increased rates of death from heart disease, and some have reported increased rates of death from cancer, liver cirrhosis, suicide, and accidents. The reasons for these associations are uncertain, and are likely to be different for different conditions. Increased rates of suicide have been reported, together with increased rates of psychological morbidity and economic disadvantage (Valdimarsdottir et al., 2003).

Management of grief

Grief is a normal response, and most people pass through it with the help of family, friends, spiritual advisers, and the rituals of mourning. In some Western societies, many people may not have links with a religion, the rituals of mourning may be attenuated, and family may not be close at hand. For these and other reasons, family doctors have an important part to play in helping the bereaved. Psychiatrists may be asked to help people with abnormal grief.

Although bereaved people have some problems in common, they also have problems that are individual. For example, a young widow with small children will have many difficulties that are not shared by an elderly widow whose adult children can support her. A mother who is grieving for a stillborn child will have special problems (discussed below). When planning management it is important to take into account the individual circumstances of the patient, as well as the general guidelines outlined below.


When counselling is appropriate, it is similar to counselling for other kinds of adjustment reaction. The bereaved person needs to talk about the loss, to express feelings of sadness, guilt, or anger, and to understand the normal course of grieving. It is helpful to forewarn a bereaved person about unusual experiences such as feeling as if the dead person is present, illusions, and hallucinations, otherwise these experiences may be alarming. Help may be needed to:

• accept that the loss is real

• work through the stages of grief

• adjust to life without the deceased.

The bereaved person may need help to progress from the first stage of denial of loss to the acceptance of reality. Viewing the dead body and putting away the dead person’s belongings help this transition, and a bereaved person should be encouraged to perform these actions. Practical problems may need to be discussed, including funeral arrangements and financial difficulties. A young widow may need help with caring for young children, and in supporting them without inhibiting her own grief excessively. As time passes, the bereaved person should be encouraged to resume social contacts, to talk to other people about the loss, to remember happy and fulfilling experiences that were shared with the deceased, and to consider positive activities that the latter would have wanted survivors to undertake. For further information about grief counseling, see Clark (2004).

Parents who are grieving for a stillborn child need special help. Previous practice has been to advise physical contact with the stillborn baby, and this is something which many parents understandably desire. However, there is growing evidence that holding the body of the baby is associated with greater risk of adverse psychological outcomes, so current advice is that parents should not view or hold the body unless they particularly wish to do so (Turton et al., 2009). Stillbirth increases the risk of subsequent relationship breakdown; lack of support by the partner at the time of the stillbirth and having held the stillborn infant increase this risk.


Drug treatment cannot remove the distress of normal grief, but it may be needed in specific circumstances. In the first stage of grief, a hypnotic or anxiolytic drug may be needed for a few days to restore sleep or to relieve any severe anxiety. In the second stage, antidepressant drugs may be beneficial if the criteria for depressive disorder are met, although such usage has not been widely evaluated in this special group. Medication may be needed for a short period in the second stage to relieve severe anxiety.

Support groups

Support groups have been developed to help recently bereaved people, particularly young widows. One such organization in the UK is known as Cruse. By sharing their experience with others who have dealt successfully with bereavement, recently bereaved people can share their grief, obtain practical advice, and discuss ways of coping (Clark, 2004).


It is not practicable, nor is there evidence that it is helpful, to provide psychotherapy for all bereaved individuals. There is some evidence that crisis intervention may be helpful for people who are at high risk of an abnormal grief reaction, although not for unselected grieving individuals. Marmar et al. (1988) studied brief dynamic psychotherapy and found it to be no more effective than a mutual support group. Similarly, Lieberman and Yalom (1992) found no significant difference in outcome between bereaved spouses treated with group psychotherapy and a control group who were not treated. For abnormal grief, dynamic psychotherapy has a clear rationale and approach, but its effectiveness has not been formally evaluated (Clark, 2004). Complicated grief treatment (CGT), which combines aspects of cognitive–behavioural therapy and interpersonal therapy, has been reported to be effective in the management of abnormal grief (Zisook and Shear, 2009).

Long-term adjustment to sexual abuse in childhood

When sexually abused, children may experience anxiety, depression, and post-traumatic stress disorder (see p. 159). These effects usually subside during childhood, but people who have been abused in childhood are more vulnerable than others to psychiatric disorder in adult life. Furthermore, sexual abuse in childhood may be followed by persistent low self-esteem and psychosexual difficulties, whether or not a psychiatric disorder develops.

Some adults who were previously unaware that they had been sexually abused in childhood suddenly recall the abuse in a vivid and disturbing way. Sometimes this recall occurs spontaneously, often when the person has encountered a reminder of the events. It may also occur during counselling or psychotherapy, at a time when childhood experiences are being discussed. Some of these recollections may be confirmed by other evidence, but many are vigorously denied by the alleged abuser, who is often one of the parents. It has been suggested that some, perhaps most, of these unconfirmed reports of abuse are not accurate memories, and that some have been induced by questions, suggestions, or interpretations from the therapists. This phenomenon has been termed false memory syndrome (Brewin, 2009).

Recovered memory and false memory

Many victims of sexual abuse, and of other severe stressful events, have partial amnesia for the most stressful parts of the experience, even though they have suffered no head injury that could lead to post-traumatic amnesia. Indeed, partial amnesia is part of the clinical picture of post-traumatic stress disorder. However, complete amnesia is less frequent and, in the view of many psychiatrists, complete amnesia for repeated stressful events followed by their recall is improbable, especially when there is no supporting evidence for the events from another source. This doubt is increased by evidence that ‘memories’ of single non-abusive childhood events can be implanted by suggestion in about 25% of subjects (Wright et al., 2006).

Evidence for the proposition that true memories can be inaccessible for many years and then be recovered comes mainly from clinical reports. These reports suggest that around 25–50% of people who report childhood sexual abuse describe long periods during which they did not remember the abuse. Furthermore, clinical studies have shown that around 20–60% of people who report childhood sexual abuse state that there were periods during their life when they could not remember that the abuse had taken place. In addition, some people have recovered memories of abuse prior to engaging in therapy, so the effect of suggestion could not have been an influence in these cases (Brewin, 2009).

In the absence of conclusive evidence about the status of memories that are recovered during counselling or psychotherapy, the clinician who is carrying out these procedures should:

• take special care not to suggest memories of sexual abuse

• consider most carefully apparent recovered memories that arise for the first time in therapy before concluding that they are true memories of actual events.

The state of the current scientific evidence indicates that practitioners should keep an open mind about the possibility and likely frequency of recovered memories of sexual abuse and, of course, of other kinds of traumatic experience.


Adults who report having experienced sexual abuse in childhood have higher rates of psychiatric disorder in adult life (Mullen et al., 1993; Kendler et al., 2000). Most studies have been retrospective and have involved women. Female psychiatric patients are more likely than healthy controls to report having experienced sexual abuse in childhood; such reporting is particularly frequent in those with somatization disorder, dissociative states, and borderline personality disorder (Sar et al., 2004). It is not clear what proportion of women who were sexually abused in childhood develop these disorders in adult life, but some make a good adjustment. In a prospective study of over 1600 abused children, Spataro et al. (2004) found that that 12.4% of the abused group subsequently received psychiatric treatment, compared with 3.6% of controls. Male children were significantly more likely to have received treatment than females. Abused subjects had higher rates of several psychiatric disorders, including a range of childhood mental disorders as well as adult mood disorders, somatoform disorders, stress disorders, personality disorder, and substance use disorders. In general, psychiatric consequences are particularly likely when the abusing person is a parent, when there are multiple abusers, or when abuse is prolonged (Steel et al., 2004).


There are three possible explanations for an association between the reporting of childhood sexual abuse and the symptoms of psychiatric disorder in adult life.

1. People with psychiatric disorder may be more likely than controls to report childhood sexual abuse, perhaps because they have been asked questions about their childhood in the course of psychiatric assessment.

2. Childhood sexual abuse may be a direct cause of vulnerability to adult psychiatric disorder.

3. Sexual abuse may be a marker of some other factor, such as disturbed relationships within the family, which is the real cause of the excess psychiatric disorder in adult life.

Each of these possible causes will now be considered in turn.

It seems unlikely that the association can be explained solely by the greater recall of sexual abuse by women who have psychiatric disorder, because community studies have also found an association between the reporting of childhood sexual abuse and the reporting of psychiatric symptoms. In addition, as noted above, there are also prospective data which indicate that sexual abuse in childhood is indeed associated with adult psychiatric disorder (Spataro et al., 2004). Finally, many patients who have been abused report that they were not in fact asked about this by mental health professionals (Read et al., 2007).

The extent to which sexual abuse itself directly causes adult psychiatric disorder is uncertain. For example, an association between childhood abuse and adult psychiatric disorder may be explained in part by disturbed relationships within the family of the abused child. Thus people who have been abused as children are more likely to report to others that their parents were uncaring or emotionally distant (Alexander and Lupfer, 1987). Mullen et al. (1993) found that, with less severe forms of abuse, the relationship between abuse and subsequent disorder could be accounted for by the family factors alone, but when abuse was severe, it had an independent and direct effect. Other studies, which have attempted to control for the various associated variables, have also concluded that abuse itself may be a direct cause of severe mental illness (Janssen et al., 2004).


The late effects of childhood sexual abuse have been treated with counselling, dynamic psychotherapy, cognitive therapy, and group treatments. The various methods have several common features.

• The general aim is to help the patient to understand the earlier experiences and the effects of these on their life, in order to improve present adjustment.

• The therapeutic relationship is used to help the patient to feel trusted, understood, and respected, and to increase their self-esteem.

• The patient is allowed to set the pace at which they talk about the experience of being abused, otherwise they may be overwhelmed by an extreme emotional response to the memories of abuse, and withdraw from treatment.

• Present problems of adjustment are identified, especially any avoidance of problems and difficulties in expressing anger. Help is given to overcome these difficulties.

• Some patients need help with psychosexual problems.

The main difference between the dynamic and cognitive–behavioural approaches is the greater emphasis given in the former to understanding the effects of the trauma on self-esteem and emotional expression, and the greater emphasis given in the latter to more precise specification of the ways in which current patterns of thinking affect present behaviour.

For a review of psychological treatment of the long-term effects of sexual abuse, see Callahan et al. (2004) and Adshead and Ferris (2007).

Further reading

Gelder MG, Andreasen NC, López-Ibor JJ Jr and Geddes JR (eds) (2009). Section 4.6: Stress-related and adjustment disorders. In: The New Oxford Textbook of Psychiatry. Oxford University Press, Oxford. (The four chapters in this section contain systematic reviews of acute stress reactions, post-traumatic stress disorder, recovered and false memories, and adjustment disorders.)

Parkes CM (2010). Bereavement. Penguin Books, London. (A comprehensive account of grief, written for the layman, but containing useful information for the professional.)

National Institute for Health and Clinical Excellence (2005). Post-Traumatic Stress Disorder (PTSD): the management of PTSD in adults and children in primary and secondary care. Clinical Guideline 26. National Institute for Health and Clinical Excellence, London. (A periodically updated comprehensive review, available at