ArticlePDF AvailableLiterature Review

Abstract and Figures

Posttraumatic stress disorder (PTSD) is a common reaction to traumatic events. Many people recover in the ensuing months, but in a significant subgroup the symptoms persist, often for years. A cognitive model of persistence of PTSD is proposed. It is suggested that PTSD becomes persistent when individuals process the trauma in a way that leads to a sense of serious, current threat. The sense of threat arises as a consequence of: (1) excessively negative appraisals of the trauma and/or ist sequelae and (2) a disturbance of autobiographical memory characterised by poor elaboration and contextualisation, strong associative memory and strong perceptual priming. Change in the negative appraisals and the trauma memory are prevented by a series of problematic behavioural and cognitive strategies. The model is consistent with the main clinical features of PTSD, helps explain several apparently puzzling phenomena and provides a framework for treatment by identifying three key targets for change. Recent studies provided preliminary support for several aspects of the model.
Content may be subject to copyright.
Invited essay
A cognitive model of posttraumatic stress disorder
Anke Ehlers*, David M. Clark
Department of Psychiatry, University of Oxford, Warneford Hospital, Oxford OX3 7JX, UK
Received 4 July 1999
Posttraumatic stress disorder (PTSD) is a common reaction to traumatic events. Many people recover
in the ensuing months, but in a signi®cant subgroup the symptoms persist, often for years. A cognitive
model of persistence of PTSD is proposed. It is suggested that PTSD becomes persistent when
individuals process the trauma in a way that leads to a sense of serious, current threat. The sense of
threat arises as a consequence of: (1) excessively negative appraisals of the trauma and/or its sequelae
and (2) a disturbance of autobiographical memory characterised by poor elaboration and
contextualisation, strong associative memory and strong perceptual priming. Change in the negative
appraisals and the trauma memory are prevented by a series of problematic behavioural and cognitive
strategies. The model is consistent with the main clinical features of PTSD, helps explain several
apparently puzzling phenomena and provides a framework for treatment by identifying three key targets
for change. Recent studies have provided preliminary support for several aspects of the model. #2000
Elsevier Science Ltd. All rights reserved.
Keywords: Posttraumatic stress disorder; PTSD; Memory; Cognitions; Cognitive behaviour therapy
Posttraumatic stress disorder (PTSD) is a common reaction to traumatic events such as
assault, disaster or severe accidents. The symptoms include repeated and unwanted
reexperiencing of the event, hyperarousal, emotional numbing and avoidance of stimuli
(including thoughts) which could serve as reminders for the event. Many people experience at
least some of these symptoms in the immediate aftermath of the traumatic event. A sizeable
proportion recover in the next few weeks or months, but in a signi®cant subgroup the
Behaviour Research and Therapy 38 (2000) 319±345
0005-7967/00/$ - see front matter #2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 00 5 - 7 9 6 7 ( 9 9 ) 0 0 1 2 3 - 0
* Corresponding author. Tel.: +44-1865-223-904; fax: +44-1865-223-908.
E-mail address: (A. Ehlers).
symptoms persist, often for years (Kessler et al., 1995; Rothbaum, Foa, Riggs, Murdock &
Walsh, 1992). It is largely the subgroup of people with persistent PTSD who seek treatment.
For these people social and occupational functioning are often severely impaired. The purpose
of this paper is to introduce a cognitive model that was designed to explain the persistence of
PTSD and to provide a framework for the cognitive-behavioural treatment of PTSD. The
model draws heavily on the writings of other theorists (Brewin, Dalgleish & Joseph, 1996;
Conway, 1997a; 1997b; Foa & Riggs, 1993; Foa & Rothbaum, 1998; Foa, Steketee &
Rothbaum, 1989; Horowitz, 1997; Jano-Bulman, 1992; Joseph, Williams & Yule, 1997;
Markowitsch, 1996; Resick & Schnicke, 1993; van der Kolk & Fisler, 1995; van der Kolk &
van der Hart, 1991). However, the model is distinct in the particular synthesis it provides.
1. A cognitive model of PTSD
1.1. Overview
When trying to conceptualise PTSD from a cognitive perspective, one is immediately
presented with a puzzle. PTSD is classi®ed as an anxiety disorder. Within cognitive models,
anxiety is a result of appraisals relating to impending threat. However, PTSD is a disorder in
which the problem is a memory for an event that has already happened. We suggest that this
apparent puzzle can be resolved by proposing that persistent PTSD occurs only if individuals
process the traumatic event and/or its sequelae in a way which produces a sense of a serious
current threat. The model proposes that two key processes lead to a sense of current threat.
1. individual dierences in the appraisal of the trauma and/or its sequelae
2. individual dierences in the nature of the memory for the event and its link to other
autobiographical memories.
Once activated, the perception of current threat is accompanied by intrusions and other
reexperiencing symptoms, symptoms of arousal, anxiety and other emotional responses. The
perceived threat also motivates a series of behavioural and cognitive responses that are
intended to reduce perceived threat and distress in the short-term, but have the consequence of
preventing cognitive change and therefore maintaining the disorder. Figure 1 summarises the
key variables in the model. Each is explained in greater detail below.
1.2. Appraisal of the trauma and/or its sequelae
It is assumed that, unlike individuals who recover naturally, individuals with persistent
PTSD are unable to see the trauma as a time-limited event that does not have global negative
implications for their future. The model proposes that these individuals are characterised by
idiosyncratic negative appraisals of the traumatic event and/or its sequelae that have the
common eect of creating a sense of serious current threat. This threat can be either external
(e.g. the world is a more dangerous place) or, very commonly, internal (e.g. a threat to one's
view of oneself as a capable/acceptable person who will be able to achieve important life goals
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345320
(see also Foa & Riggs, 1993; Jano-Bulman, 1992; Joseph et al., 1997; Meichenbaum, 1997;
Resick & Schnicke, 1993)). Examples are given in Table 1.
1.2.1. Appraisal of the traumatic event
Several types of appraisal of the traumatic event can produce a sense of current threat. First,
individuals may overgeneralise from the event and as a consequence perceive a range of normal
activities as more dangerous than they really are. They may exaggerate the probability of
further catastrophic events in general or take the fact that the trauma happened to them, as
opposed to other people, as evidence for appraisals such as ``I attract disaster'' or ``bad things
always happen to me''. Such appraisals generate not only situational fear but also avoidance
which maintains the overgeneralised fear. Common examples include avoiding driving after
experiencing a road trac accident because of an unrealistic belief about the likelihood of
future accidents or severely restricting one's social life after a sexual assault for a similar
Second, appraisals of the way one felt or behaved during the event can have long-term
threatening implications. For example, a woman who experienced sexual arousal during a
particularly protracted rape interpreted her response as a sign that she had secret desires that
were repulsive to her. Similarly, another woman who was raped by an acquaintance interpreted
Fig. 1. A cognitive model of PTSD.
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 321
her inability to spot that this was likely to happen as a sign that she was much less capable of
`reading' other people than she thought and that she should therefore abandon her cherished
plans for a career in clinical psychology.
1.2.2. Appraisals of trauma sequelae
A variety of idiosyncratic, negative appraisals of the sequelae of the traumatic event can
produce a sense of current threat and contribute to persistent PTSD. These include:
interpretation of one's initial PTSD symptoms, interpretation of other people's reactions in the
aftermath of the event and appraisal of the consequences that the trauma has in other life
domains (e.g. physical consequences such as pain and ®nancial or professional consequences).
Symptoms such as intrusive recollections and ¯ashbacks, irritability and mood swings, lack
Table 1
Examples of idiosyncratic, negative appraisals leading to sense of current threat in persistent PTSD
What is appraised? Negative appraisal
Fact that trauma happened ``Nowhere is safe''
``The next disaster will strike soon''
Trauma happened to me ``I attract disaster''
``Others can see that I am a victim''
Behaviour/emotions during trauma ``I deserve that bad things happen to me''
``I cannot cope with stress''
Initial PTSD symptoms
Irritability, anger outbursts ``My personality has changed for the worse''
``My marriage will break up''
``I can't trust myself with my own children''
Emotional numbing ``I'm dead inside'',
``I'll never be able to relate to people again''.
Flashbacks, intrusive recollections and nightmares ``I'm going mad'', ``I'll never get over this''.
Diculty concentrating ``My brain has been damaged'', ``I'll lose my job''.
Other people's reactions after trauma
Positive responses ``They think I am too weak to cope on my own''
``I am unable to feel close to anyone''
Negative responses ``Nobody is there for me''
``I cannot rely on other people''
Other consequences of trauma
Physical consequences ``My body is ruined''
``I will never be able to lead a normal life again''
Loss of job, money etc. ``I will lose my children''
``I will be homeless''
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345322
of concentration and numbing are common reactions shortly after a traumatic event. If
individuals do not see these symptoms as a normal part of the recovery process, they may
interpret them as indications that they have permanently changed for the worse or as
indicators of a threat to their physical or mental well being (see also Ehlers & Steil, 1995; Foa
& Riggs, 1993; Foa & Rothbaum, 1998; Jones & Barlow, 1990). Table 1 gives several examples
of negative appraisals of initial PTSD symptoms. Such appraisals maintain PTSD by directly
producing negative emotions (e.g. anxiety, depression or anger) and by encouraging individuals
to engage in dysfunctional coping strategies that have the paradoxical eect of enhancing
PTSD symptoms. For example, individuals who believe that intrusive recollections mean they
are losing control of their mind are likely to try hard to push such recollections out of their
mind. Unfortunately, active thought suppression of this type often makes the thought more
likely to come to mind (Wegner, 1989).
Other people, including family and close friends, are often uncertain about how they should
respond to a trauma victim and may avoid talking about the event in order not to distress the
victim. This `consideration' can be interpreted as a sign that others do not care, or, worse still,
that they think the event was partly the victim's fault. Such interpretations are likely to directly
produce some of the symptoms of PTSD (estrangement from others and social withdrawal)
and are also likely to prevent victims from discussing the trauma with others, hence reducing
the opportunity for therapeutic reliving (see below) and for feedback from others that might
help correct excessively negative views about the meaning of the event. Of course some people
are also objectively uncaring, rejecting or critical of victims after a traumatic event. If
traumatised individuals consider these people's views important, they may interpret such
reactions as a sign that they are to blame for the event, that they are unworthy, that they are
unlikeable or that they will not be able to have close relationships with others.
Traumatic events can have negative long-term eects on many areas of life, including the
individual's physical health, appearance, vocational and ®nancial situation. These can be
interpreted as a sign of a permanent negative change of one's life for the worse or as a sign
that worse is still to come.
1.2.3. Appraisals and emotional responses
The nature of predominant emotional responses in persistent PTSD depends on the
particular appraisals (see Beck, 1976). Appraisals concerning perceived danger lead to fear (e.g.
``Nowhere is safe''), appraisals concerning others violating personal rules and unfairness lead
to anger (e.g. ``Others have not treated me fairly''), appraisals concerning one's responsibility
for the traumatic event or its outcome lead to guilt (e.g. ``It was my fault''), appraisals
concerning one's violation of important internal standards lead to shame (e.g. ``I did something
despicable'') and appraisals concerning perceived loss lead to sadness (e.g. ``My life will never
be the same again''). Most patients with persistent PTSD experience a range of negative
emotions. This is partly because dierent appraisals are activated at dierent times and partly
because the degree of conviction varies from time to time. For example, the possibility that a
loss may occur tends to be associated with anxiety whereas perceived certainty of a loss tends
is associated with depression.
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 323
1.3. Memory for the traumatic event
The nature of trauma memory and its relationship to unwanted recollections is another
puzzle of persistent PTSD. On the one hand, patients often have diculty in intentionally
retrieving a complete memory of the traumatic event. Their intentional recall is fragmented and
poorly organized, details may be missing and they have diculty recalling the exact temporal
order of events (Foa & Riggs, 1993; Foa, Molnar & Cashman, 1995; van der Kolk & Fisler,
1995; Koss, Figueredo, Bell, Tharan & Tromp, 1996; Amir, Staord, Freshman & Foa, 1998).
On the other hand, patients report a high frequency of involuntarily triggered intrusive
memories involving reexperiencing aspects of the event in a very vivid and emotional way.
Models of PTSD need to explain this apparent discrepancy between diculties in intentional
recall and easily triggered reexperiencing of the event. In addition, the involuntary
reexperiencing has a number of important characteristics that need to be explained. These
characteristics will be described ®rst. We will then go on to outline a possible explanation for
the memory disturbance.
.Reexperiencing mainly consists of sensory impressions, rather than thoughts. The
impressions can involve all modalities including physical sensations, but are predominantly
visual (see Ehlers & Steil, 1995; van der Kolk & Fisler, 1995). For example, a man who
experienced a head-on car crash at night kept seeing headlights coming towards him.
.The sensory impressions are experienced as if they were happening right now rather than
being memories from the past and the emotions (including physical reactions and motor
responses) accompanying them are the same as those experienced at the time (`original'
emotions). They lack the awareness of remembering that usually characterises
autobiographical memories (see also Foa & Rothbaum, 1998; and Brewin et al.'s, 1996,
concept of situationally accessible memories). Reemtsma (1997) who was kidnapped and held
hostage in a cellar for a month provides a good illustration. After the kidnappers set him
free, the terror he experienced in the cellar kept haunting him. He describes this experience
as ``being back in the cellar''. Note that he does not say it was like being back in the cellar.
.The original emotions and sensory impressions are reexperienced even if the individual later
(i.e. at another time during the event or afterwards) acquired new information that
contradicted the original impression or if he/she knows that these impressions did not turn
out to be true. For example, a patient whose father committed suicide by shooting himself,
kept reexperiencing a panicky urge to ®nd his father and the feeling of responsibility for
rescuing him that he had when he discovered the suicide note. At the time, he erroneously
thought that his father had taken sleeping tablets and could be saved if he acted quickly
.``Aect without recollection''. Individuals with PTSD sometimes reexperience physiological
sensations or emotions that were associated with the traumatic event without a recollection
of the event (lack of source information, see also Schacter, Norman & Koutstaal, 1997). For
example, a rape victim noticed that she was feeling extremely anxious while talking to a
female friend in a restaurant and only subsequently realised that the feeling was probably
triggered by the presence of a man on another table who bore some physical similarity with
the rapist.
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345324
.The involuntary reexperiencing of the traumatic event is triggered by a wide range of stimuli
and situations. Many of the trigger stimuli are cues that do not have a strong semantic
relationship to the traumatic event, but instead are simply cues that were temporally
associated with the event. Common examples are physical cues similar to those present
shortly before or during the traumatic event (e.g. the shape of a person, spatial cues, smells,
a pattern of light, particular phrases said in a certain tone of voice), similar emotional states
(e.g. feeling helpless or trapped) or other similar internal cues (e.g. touch on a certain part of
the body, proprioceptive feedback from one's own movements or posture).
It is proposed that the intrusion characteristics and the pattern of retrieval that characterises
persistent PTSD (poor intentional recall, vivid unintentional reexperiencing with `here and
now' quality) is due to the way the trauma is encoded and laid down in memory.
1.3.1. Poor elaboration and incorporation into autobiographical memory base
There are two routes to the retrieval of autobiographical information (see also Brewin et al.,
1996). First, through higher-order meaning-based retrieval strategies (e.g. remembering the ®rst
day at school). Second, through direct triggering by stimuli that were associated with the event
(e.g. particular smells or a piece of music). Much of the normal processing of autobiographical
memories appears to have the function of reducing the ease with which memories of past
experiences are unintentionally retrieved while we are engaging in everyday tasks.
Autobiographical events are usually incorporated into an autobiographical memory knowledge
base that is organized by themes and personal time periods (Conway & Pleydell-Pearce, 1997).
This type of elaboration enhances the ®rst retrieval route and inhibits the second (see Conway,
1997b; Markowitsch, 1995) and has the eect that when an autobiographical memory enters
consciousness, it comprises both speci®c information about the event and general information
about the lifetime period that the event took place in and abstracted information about the
type of event in general (e.g. school days in general).
We propose that in persistent PTSD one of the main problems is that the trauma memory is
poorly elaborated and inadequately integrated into its context in time, place, subsequent and
previous information and other autobiographical memories (see also Siegel, 1995). This
explains problematic intentional recall (weak semantic route to retrieval), the `here and now'
quality (no context in time, hence the perception of current threat), the absence of links to
subsequent information (e.g. ``I did not die'') and the easy triggering by physically similar cues.
1.3.2. Strong S±S and S±R associations
We propose that a further problem in persistent PTSD is that S±S and S±R associations are
particularly strong for traumatic material. This makes triggering of memories of the event and/
or emotional responses by associated stimuli even more likely (see also Conway, 1997a; Foa et
al., 1989; Keane, Zimering & Caddell, 1985; Charney, Deutch, Krystal, Southwick & Davis,
1993). A good illustration of how S±S associative learning leads to involuntary reexperiencing
is given by Reemtsma (1997). His most distressing intrusion after his release from the cellar
was hearing a knocking sound, and he experienced great distress with this intrusion. His
kidnappers had knocked at the door of the cellar when bringing him food, water, etc. When
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 325
they knocked, he had to lie down immediately with his face to the ¯oor and make sure he did
not see them, knowing he would be killed if he did. He describes that initially the intrusive
knocking sound appeared to come out of the blue, but that he gradually became aware that
this intrusion was often triggered by hearing footsteps. During his captivity, he had heard
footsteps approaching before the kidnappers knocked at the door. The sound of footsteps had
become associated with the sound of knocking.
Two aspects of S±S and S±R associative learning are of interest in explaining the persistence
of PTSD. First, this form of learning helps the organism in making predictions (including
those that operate outside awareness) about what will happen next. It appears that in PTSD
stimuli that were present shortly before or during the traumatic event become
associated with the default prediction of severe danger to self. Second, retrieval from
associative memory is cue-driven and unintentional so that the individual may not always be
aware of the triggers for reexperiencing (as in Reemtsma's example) and may not be aware that
his/her emotional reaction is due to activation of the trauma memory (aect without
. Failure to spot the origin of the reexperiencing symptoms makes it dicult for
the patient to learn that there is no present danger when exposed to the triggers
1.3.3. Strong perceptual priming
We propose that there is particularly strong perceptual priming (a form of implicit memory)
for stimuli that were temporally associated with the traumatic event, i.e. there is a reduced
perceptual threshold for these stimuli. As a consequence of the reduced perceptual threshold,
cues that were associated with the trauma and that consequently can directly trigger the
trauma memory are more likely to be noticed. As implicit memory traces are not well
discriminated from other memory traces (Baddeley, 1997), vague physical similarity would be
sucient in perceiving stimuli as similar to those occurring in the traumatic situation (poor
stimulus discrimination) and thus triggering reexperiencing symptoms, even if the context in
which the stimulus con®guration is observed is very dierent. For example, a patient who had
been involved in a car crash at night noticed that a patch of bright sunlight on his lawn
triggered vivid intrusions of headlights coming towards him.
1.4. Relationship between the nature of trauma memory and trauma appraisals
There is a reciprocal relationship between the nature of the trauma memory and the
appraisals of the trauma/its sequelae. When individuals with persistent PTSD recall the
traumatic event, their recall is biased by their appraisals and they selectively retrieve
With Rescorla (1988) we assume that this is an `intelligent' process in that stimuli with a high information value
(in terms of predicting the occurrence of the traumatic event) are particularly likely to become associated with
danger. Note that the stimuli may not have a meaningful relationship with danger.
In line with the hypothesis that these reexperiencing phenomena re¯ect strong S±S and S±R associative learning
and a relative de®cit in memory elaboration, Bechara et al. (1995) have demonstrated a dissociation between con-
ditioning and declarative memory associated with amygdala and hippocampal functions.
This corresponds to LeDoux's ®nding from animal research that involvement of the cortex is necessary to
unlearn conditioned fear responses (LeDoux, 1992).
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345326
information that is consistent with these appraisals. For example, a patient who thought that
the trauma (an accident) showed that nobody cared about her, recalled unfriendly responses of
nurses in hospital, but did not recall that several people had tried to help her after the
accident. Such selective retrieval prevents individuals from remembering aspects of the
traumatic event that contradict their appraisals and thus prevents change in the appraisals.
When the patient remembered during imaginal reliving that others were trying to help after the
accident, her belief that nobody cared about her decreased.
On the other hand, inability to remember details of the trauma can be appraised by
individuals in a way that maintains the sense of current threat, for example, that the memory
problem means something is seriously wrong with them (e.g. brain damage) or that something
even worse must have happened during the trauma that would be unbearable if they knew
what it was. Inability to remember the exact order of events can contribute to the erroneous
appraisal of being responsible for the event.
Similarly, the `here and now' quality of the emotions that are associated with the trauma
memory can contribute to problematic appraisals. For example, many people feel extremely
lonely during a traumatic event and reexperiencing these feelings in the company of signi®cant
others may be interpreted as a sign that they are unable to relate to other people or that their
relationships with others have permanently changed for the worse.
Furthermore, it is proposed that in those people with persistent PTSD for whom the
traumatic event has seriously threatened their view of themselves (e.g. as worthy or capable),
the general organisation of their autobiographical memory knowledge base may be disturbed
(an extreme case of a patient who developed complete retrograde amnesia for the past 6 years
after a traumatic event is described by Markowitsch et al. (1998)). Such people seem unable to
reorganise their previous and subsequent experiences in a way which produces a stable view of
themselves and the context they live in (see Conway's idea that autobiographical knowledge
grounds the self (Conway, 1997a; Conway & Pleydell-Pearce, 1997)). This will produce a sense
of disorientation and will also have the eect that their retrieval from memory will be less
®ltered by current context and more cue-driven than the perceptions of other people with a
strong sense of self in context. This is comparable to a person who has moved to a new town
and keeps `seeing' people from his previous home town by responding to vague physical
similarity until he establishes a clear awareness of himself in a new environment. The
disorganised autobiographical memory will therefore make cue-driven recollections of the
traumatic event/aect more likely.
1.5. Maladaptive behavioural strategies and cognitive processing styles
When patients with persistent PTSD perceive a serious current threat and the accompanying
symptoms, they try to control the threat and symptoms by a range of strategies. The strategy
selected is meaningfully linked with the individual's appraisals of the trauma and/or its
sequelae and their general beliefs about how best to deal with the trauma. Further examples
This may be re¯ected in the ®ndings of poor retrieval of speci®c autobiographical memories in PTSD (Kuyken &
Brewin, 1995; McNally et al., 1995).
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 327
are given in Table 2. The strategies intended to control the threat/symptoms are maladaptive
because they maintain PTSD by three mechanisms:
1. Directly producing PTSD symptoms,
2. Preventing change in negative appraisals of the trauma and/or its sequelae,
3. Preventing change in the nature of the trauma memory.
An example of a maladaptive cognitive strategy that increases PTSD symptoms directly is
thought suppression. If patients try hard to push thoughts about the trauma out of their mind,
this will increase the frequency of unwanted intrusive recollection. Another common example is
that behaviours used to control some of the PTSD symptoms may increase others, e.g. attempts
to prevent nightmares by going to bed very late or getting up very early may increase
symptoms of poor concentration, irritability and alienation. Selective attention to threat cues
another example of a cognitive process that may increase the frequency of intrusions and
trauma-related emotions.
Among the strategies that prevent a change in the appraisal of the traumatic event or its
sequelae are safety behaviours. These are actions individuals take to prevent or minimise
anticipated further catastrophes (Salkovskis, 1996). Safety behaviours prevent discon®rmation
of the belief that the feared catastrophe will occur if one does not engage in preventative
action. For example, individuals may be extremely vigilant for possibly dangerous situations
while driving in order to decrease the probability of another accident. Individuals who were
assaulted in their homes may always sleep with a knife next to their bed in order to minimise
the risk of being killed by another intruder.
Among the maladaptive strategies that prevent a change in the nature of the trauma memory
is actively trying not to think about the event. Individuals with persistent PTSD try to keep
their mind constantly occupied with other things or they try to think about the event in a non-
emotional way (like giving a report to the police or a journalistic description), leaving out the
parts with the largest emotional impact. These eorts can take elaborate forms. For example, a
lorry driver who had been involved in a fatal accident kept occupying his mind with sexual
fantasies when at work to prevent memories of the accident from popping back into his mind.
Another patient spent hours cleaning her house to prevent being overwhelmed by memories.
Eorts to not think about the event prevent individuals from elaborating the trauma memory
and linking their experience with its context in time, space, previous and subsequent
information and other autobiographical memories. They also prevent changes in appraisals
about what would happen if they thought about the trauma (e.g. ``I will go mad'').
Similarly, avoidance of reminders of the trauma maintains PTSD by preventing both a change
in the problematic appraisals (e.g. ``If I encounter ..., the trauma will happen again'', see also
Table 2) and a change in the nature of the memory. Avoidance of the site of the trauma
commonly prevents correction of appraisals about how the event could have been avoided. As
reminders of the trauma often provide retrieval cues for inaccessible details, avoidance of these
We talk about the dysfunctional behaviours and cognitive processes as strategies, but we do not assume that they
always have an intentional quality. They may be performed in a habitual or re¯exive fashion. For example, selective
attention to threat and dissociation probably includes automatic as well as strategic responses. The former may rep-
resent part of the trauma memory that can be automatically triggered when reminders are present.
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345328
Table 2
Examples of appraisals with associated dysfunctional behavioural and cognitive strategies
Appraisal Dysfunctional strategies
If I think about the trauma
...I will go mad
...I will fall apart
try hard not think about the trauma; keep mind
occupied all the time; control feelings; drink
alcohol/ take drugs
...I will lose control and hurt someone
...I will have a heart attack
...I will seriously damage my health
If I do not control my feelings tightly
...I will not be able to work and lose my job
numb emotions; avoid anything that could cause
negative or positive feelings
...I will lose my temper and oend people
If I do not ®nd out how this event could have been
ruminate about how event could have been
...something similar will happen again
If I do not ®nd a way to punish the assailant he will have
won and l will not be a proper man any longer
ruminate about how to get even with assailant
If I go to the site of the event, avoid site of the event
If I wear the same clothes again, avoid wearing similar clothes
...I will have another accident
...I will have a nervous breakdown
If I do not take extra precaution
...I will be attacked again
carry weapon; vigilant for dangerous people; avoid
crowded places; make sure to stay close to exit
If I do not check the rear mirrors keep checking mirrors
...someone will drive into my car again
If I make plans (such as for a holiday) do not make any plans for the future
...the next awful thing is going to happen
If I see my friends avoid seeing friends
...they will ask me about the trauma and they will think
that I am pathetic because I am still so upset
If I do things that I used to enjoy give up pleasant activities
...I will be punished again
...I will be reminded of the trauma and will not be able to
...I will be overwhelmed by emotion
If I show my face
...people will be disgusted because of my scars
avoid other people; cover face with hands; heavy
make-up; look down
(continued on next page)
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 329
cues also interferes with the formation of a more elaborate trauma memory that links the
experience to its context. Similarly, use of alcohol or medication to control anxiety will prevent
a change in interpretations such as ``I am going to lose control when I let my feelings come''
and will also interfere with a change in the nature of the memory. Furthermore, it is common
for people with persistent PTSD to give up or avoid activities that were important to them
before the traumatic event, for example sports, hobbies or socialising. This prevents a change
in their appraisals, e.g. that the trauma has made them a dierent person or that other people
will respond negatively if they knew about the trauma and prevents them from reorganising
their autobiographical memory knowledge base in a way that creates a continuous view of the
Another common example of a maladaptive cognitive processing style is rumination about
the trauma and its consequences, for example about how it could have been prevented or
about how justice/revenge can be achieved. At this stage, it is unclear what exactly the
mechanisms are by which rumination maintains PTSD. It probably strengthens problematic
appraisals of the trauma (e.g. ``The trauma has ruined my life'') and is probably similar to
cognitive avoidance in interfering with the formation of a more complete trauma memory
because it focusses on `what if ...' questions rather than on the experience of the trauma as it
actually happened. Finally, it may also directly increase feelings of nervous tension, dysphoria
or hopelessness and, because it provides internal retrieval cues, intrusive memories of the
traumatic event.
Note that the present model assumes that dierent mechanisms underlie rumination and
reexperiencing symptoms (see also Joseph et al., 1997). Clinical descriptions of intrusive
thoughts in PTSD have not always made this distinction. Rumination is thought to be
driven by problematic appraisals whereas de®cits in the trauma memory are seen as the
cause of persistent reexperiencing symptoms. However, reexperiencing may lead to
rumination and rumination may provide internal cues that trigger reexperiencing
Dissociation when reminded of the trauma is an as yet poorly understood cognitive response
that interferes with recovery. We speculate that the derealisation, depersonalisation and
emotional numbing experienced during dissociation may impede the elaboration of the trauma
memory and its integration into the autobiographical memory knowledge base (see also Foa &
Hearst-Ikeda, 1996).
Table 2 (continued )
Appraisal Dysfunctional strategies
If I go to sleep stay up until very late
...I will have nightmares
...I will not notice intruders
If I have more stress avoid anything that could be stressful
...I will have a heart attack
...I will have a nervous break-down
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345330
1.6. Cognitive processing during trauma
The two processes that lead to a sense of serious current threat in PTSD (appraisals of the
trauma/its sequelae and the nature of the trauma memory) are themselves in¯uenced by the
type of cognitive processing during the traumatic event.
1.6.1. In¯uences on appraisal
An example of thought processes during the trauma that in¯uence subsequent appraisals is
mental defeat. Ehlers et al. (1998a); Dunmore, Clark and Ehlers (1997, 1998, 1999) and Ehlers,
Maercker and Boos (in press) identi®ed mental defeat as a correlate of chronic PTSD and poor
response to exposure treatment. Mental defeat refers to the perceived loss of all psychological
autonomy, accompanied by the sense of not being human any longer. Patients who experienced
mental defeat are more likely than other victims to interpret the trauma as evidence for a
negative view of themselves, for example, that they are unable to cope with stress, that they are
not a worthy person or that they are permanently damaged by the trauma.
1.6.2. In¯uences on memory
The nature of the trauma memory depends on the quality of processing at encoding (see also
Krystal, Bennett, Bremner, Southwick & Charney, 1995; Schacter et al., 1997; Siegel, 1995). An
important dimension of encoding is conceptual vs. data-driven processing (Roediger, 1990).
Some trauma victims describe that their thinking was extraordinarily clear and that they kept
analysing the situation whereas others report confusion and overwhelming sensory impressions.
It is suggested here that the latter group is more likely to suer from persistent PTSD because
the degree of conceptual processing (i.e. processing the meaning of the situation, processing it
in an organized way and placing it into context) during a traumatic event determines the
nature of the memory and thus the ability to intentionally retrieve information from this
memory. If the individual lacks conceptual processing and engages mainly in data-driven
processing (i.e. processing the sensory impressions), then the trauma memory will be relatively
dicult to retrieve intentionally and at the same time there will be relatively strong perceptual
priming for accompanying stimuli, in line with the results of experimental cognitive psychology
(reviewed by Roediger (1990)). The resulting memory trace will be poorly discriminated from
other memory traces (Baddeley, 1997), thus impairing stimulus discrimination between stimuli
present during the trauma and harmless stimuli that bear some similarity to these.
Besides the role of conceptual vs. data-driven processing, the unorganised memories
observed in persistent PTSD may in part result from an inability to establish a self-referential
perspective while experiencing the trauma that can be integrated into the continuum of other
autobiographic memories in time (see Wheeler, Stuss & Tulving, 1997).
It has been suggested that dissociation during trauma explains the fragmentation of
traumatic memories (Spiegel, 1991; van der Kolk & Fisler, 1995). Dissociation is a complex
concept that has several dierent components. Some of these may overlap with the concepts of
conceptual vs. data-driven processing and lack of self-referential perspective when encoding.
Emotional numbing may be a further factor that interferes with the formation of an organized
memory of the traumatic event (see Foa & Hearst-Ikeda, 1996).
Another problem at encoding that may explain some features of the trauma memory stems
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 331
from observations that propositions are stored in long-term memory with a default `true' value
(Conway, 1997b). During a traumatic event, individuals may not have enough cognitive
capacity to decide that some very threatening aspects of the trauma are not true. For example,
a rape victim remained convinced that she was unattractive because the rapist had repeatedly
told her she was ugly. The extreme distress and anxiety she experienced during the rape made
it impossible for her to appreciate that these words were untrue and instead were simply a
strategy that the rapist used to manipulate and humiliate her. She thus encoded his statements
as true and the appraisal that she was unattractive continued to pose a current threat to her.
1.7. Trauma characteristics, previous experiences and beliefs, current state
The model takes into account several background factors that are likely to in¯uence:
cognitive processing during the traumatic event, the nature of the trauma memory, individuals'
appraisals of the trauma/its sequelae and the strategies they use to control the perceived threat/
symptoms (see Fig. 1). These background factors are considered neither necessary nor sucient
factors in the etiology of persistent PTSD and the examples given below are meant as
illustrations rather than an exhaustive list.
Cognitive processing during a traumatic event will depend on a number of factors.
Characteristics of the trauma such as duration and predictability may exert an in¯uence. For
example, a road trac accident in which one is suddenly hit from the back is more dicult to
conceptually process than an accident that one can see coming. Another example is that mental
defeat is unlikely to be experienced during assault of very short duration (Ehlers et al., 1998a).
Previous experience of trauma and coping styles used during these events may play a role. For
example, victims of childhood sexual abuse may engage in little conceptual processing during a
renewed trauma because the trauma reactivates memories of the abuse during which they
primarily engaged in data-driven processing. Young children are particularly likely to engage in
data-driven processing during abuse because it is dicult for them to conceptualise what is
happening to them. Low intellectual ability may be related to a less conceptual and more data-
driven processing (see McNally & Shin's, 1995, ®ndings of an association of low intelligence
and PTSD). Prior beliefs may play a role. For example, individuals who believe that no one
could ever harm them may ®nd it hard to understand what is going on when they are
assaulted. State factors such as alcohol consumption, general exertion, degree of arousal and
fear may in¯uence the ability to process the situation in a conceptual and organized way (see
also Foa & Riggs, 1993; van der Kolk & Fisler, 1995). The impact of high arousal and fear on
trauma memory probably includes cognitive and biological pathways. For example, very high
cortisol levels during extreme stress may interfere with the encoding of the memory for the
event, thus impairing intentional recall (see also Newcomer et al., 1999).
Appraisals of the trauma and its sequelae will also in part be in¯uenced by characteristics of
the event and its sequelae. For example, if individuals perceived no control at all over the
traumatic situation, they may interpret this situational lack of control as evidence that they
have little control over their lives in general. Traumas that leave the individual with permanent
health problems are more likely to lead to appraisals such as ``My life is ruined'' than traumas
which in¯icted reversible injuries. The quality of other people's reactions in the aftermath of
the trauma (social support versus negative reactions) in¯uences the probability of appraisals
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345332
such as ``Nobody cares about me''. Prior beliefs will be important in that trauma victims with
prior negative beliefs about themselves may see the trauma as a con®rmation of these beliefs
and those with extremely positive beliefs may ®nd that the trauma shatters their trust in
themselves or the world (see Foa & Riggs, 1993; Jano-Bulman, 1992; Resick & Schnicke,
1993). Another example of the in¯uence of prior beliefs is that people who think that they
should always be in control of their emotions and thoughts may be especially likely to interpret
the intrusive reexperiencing symptoms as a sign that they are falling apart, going mad or have
a brain injury. Prior experiences can exert an in¯uence in that previous negative experiences
and traumas may be linked with the renewed trauma and may give it additional negative
meaning. For example, a victim of child sexual abuse who is raped as an adult may interpret
the rape as showing that she is the type of person who deserves no better or brings out the
worst in other people. A renewed trauma may also act as a powerful cue for memories of
earlier trauma if some of its sensory components overlap, so that it reactivates some of the
emotional responses to this earlier experience. For example, a patient who had a relatively
minor car accident was reminded by the sound of the impact of an earlier accident in which his
mother was killed. He blamed himself for this earlier accident, but had overcome his initial
distress and had managed not to think about it for many years. The second accident brought
back intrusive memories of the ®rst accident and strong feelings of guilt and the patient
developed persistent PTSD.
Cognitive and behavioural strategies used to control PTSD symptoms and current threat are
likely to be in¯uenced by prior experiences and beliefs. For example, a person who thinks that
people with emotional problems are inferior is more likely to use thought suppression when
distressing intrusive recollections of the trauma occur than other people who do not hold this
belief. The same would be true for someone who believes that there is only so much distress
that an individual can tolerate before going mad or suering ill health. People who were
criticised or ridiculed when showing fear or sadness in their childhood may try to numb their
emotions and avoid talking to others about the traumatic event.
2. Features of PTSD explained by the model
2.1. Delayed onset of PTSD
So far we have presented PTSD as a syndrome characterised by common initial symptoms,
which persist in some individuals. While this is generally correct, there are individuals with
persistent PTSD who report that they experienced few or even no symptoms in the ®rst few
weeks or months after the traumatic event and that the onset of PTSD did not occur till
months or even years after the trauma. How does the model deal with delayed onset cases? In
general, we assume that the delay occurs either because some later event gives the original
trauma or its sequelae a much more threatening meaning (see also the phenomenon of UCS
revaluation, Davey, 1989) or because some of the stimuli that are particularly potent reminders
of the traumatic event were not available until some time afterwards. A common example for
the change in meaning process are individuals who witness horri®c events as part of their
profession (e.g. ambulance workers, police). They may experience delayed PTSD if the events
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 333
become relevant to their personal lives (see Clohessy & Ehlers, in press). For example, they
may start reexperiencing removing the bodies of children from an accident site when their
children reach the same age and shape as the dead children. A common example for the
exposure to potent reminders process are individuals who are hospitalised for severe injuries
after motor vehicle accidents. While in hospital, these individuals do not encounter reminders
such as cars or the site of the accident and their minds are usually focussed on the physical
injuries and medical procedures, rather than the event which caused them.
2.2. Anniversary reactions
Many people with persistent PTSD experience aggravation of symptoms around the
anniversary of the event. These may be explained by a combination of the presence of
reminders and appraisal of the PTSD symptoms. Around anniversaries, patients are confronted
with many external reminders (such as weather and light conditions or other people asking
about it) and they also generate internal retrieval cues by dwelling on what their lives were like
before the traumatic event and about their feelings and experiences on the day, before the
traumatic event happened. Furthermore, anniversaries often are taken as landmarks for
negative appraisals of PTSD symptoms such as ``I am inadequate because I am still not over
it''. Such appraisals activate strategies (e.g. thought suppression) which prolong/intensify the
2.3. Frozen in time
Patients with persistent PTSD say that they feel locked into the past (see also Herman,
1992). They seem unable to resume their former life or to start a new life. This is illustrated by
McNally's, Lasko, Macklin and Pitman's (1995) description of Vietnam veterans who decades
after the war still wear their uniform and other regalia. Patients with chronic PTSD feel
disconnected from their former self and their life goals.
This state of being `frozen in time' has three sources. First, it is related to appraisals of the
trauma/its sequelae. For example, patients may think that they are permanently changed for
the worse by the trauma and thus `life will never be the same again'. They may also believe
that their former life goals are unimportant following such an extreme experience or irrelevant
because another catastrophe is going to happen soon. Second, continually reexperiencing
sensations and emotions they had at the time of the trauma in their original form, disconnects
them from current reality. Third, giving up or avoiding activities that were important to the
person before the traumatic event contributes to the sense that time has stood still at the point
of the traumatic event.
2.4. Sense of impending doom
Intrusive memories of the traumatic event are often accompanied by a sense of `worse is to
come', comparable to anticipatory anxiety, that motivates suppression of the memories. At ®rst
sight, this appears paradoxical as the individual obviously knows what the outcome of the
traumatic event was. The model explains the sense of worse is to come by the nature of the
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345334
trauma memory, i.e. sensory information and emotions are retrieved from the memory without
the time-perspective of `remembered' emotions, thus leading to the perception of future threat.
Furthermore, we have suggested elsewhere that intrusive memories are about `warning signals'
that during the traumatic event actually predicted the occurrence of the worst moments
(Ehlers, Hackmann, Steil, Clohessy & Wenninger, 1999). In addition, the poor ability to
retrieve details or order of events during the trauma together with the intrusive nature of the
memories, may be interpreted by individuals as indicating that something even worse happened
that they will ®nd unbearable or that it will be unbearable to face all the horrible events
2.5. No bene®t from talking/thinking about the trauma
People with persistent PTSD often report that they constantly think and talk about the
trauma, but that this has not helped them to feel any dierent. It is proposed that this is
because of the way they think and talk about the event. First, thinking in these cases often
takes the form of rumination about `what if...' questions rather than going over in one's mind
about what exactly happened and how one felt and thought during the event. Second, talking
is often done in a nonemotional way, as if giving a report to the police or aspects that the
individual ®nds most distressing are left out. This prevents proper access to the meaning of the
event and its contextualisation (see also Foa & Kozak, 1986; Pennebaker, 1989).
3. Treatment implications
When people talk about recovering from a traumatic experience, they often use the
metaphor ``I have put it in the past''. The current model suggests that in persistent PTSD,
putting the trauma into the past requires change in three areas.
.The trauma memory needs to be elaborated and integrated into the context of the
individual's preceding and subsequent experience in order to reduce intrusive reexperiencing.
.Problematic appraisals of the trauma and/or its sequelae that maintain the sense of current
threat need to be modi®ed.
.Dysfunctional behavioural and cognitive strategies that prevent memory elaboration,
exacerbate symptoms or hinder reassessment of problematic appraisals need to be dropped.
A wide range of cognitive-behavioural interventions could be used to achieve change in these
three areas (see, for example, Foa & Rothbaum (1998), Joseph et al. (1997), Meichenbaum
(1997); Resick & Schnicke (1993)). Future research will identify which interventions are most
Below we describe the procedures that the Oxford Cognitive Therapy Trauma Group
found particularly helpful in pilot work aimed at devising an ecient CBT intervention. Some
of the procedures utilise techniques that are already well-known in the ®eld. For these
David M. Clark, Anke Ehlers, Melanie Fennell, Ann Hackmann and Freda McManus.
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 335
techniques, we mainly focus on how the techniques should be implemented to maximise change
in the three target areas.
3.1. Assessment
A key aim of the assessment interview is to identify the main cognitive themes that will be
addressed in therapy. Completion of the Post-traumatic Cognitions Inventory (PTCI; Foa,
Ehlers, Clark, Tolin & Orsillo, in press), which covers a wide range of potentially problematic
appraisals, can be helpful. In addition, patients are asked to look back at the event and
consider what are the worst things about it/the most painful moments. In both the assessment
interview and subsequent therapy sessions, parts of the memory that currently elicit
particularly strong distress (`hot spots') are explored to identify meanings, as are intrusive
images and moments when the patient dissociates or withdraws from processing. The nature of
the predominant emotions (e.g. guilt, anger, shame, sadness or fear) is also an invaluable clue
to cognitive themes. To identify problematic appraisals of the trauma sequelae, it is useful to
ask what has been most distressing/dicult since the event and to explore patients' beliefs
about their symptoms, their future and other people's behaviour. In delayed onset cases, the
therapist tries to identify posttrauma events that may have changed the meaning of the original
trauma or its sequelae. To identify problematic behavioural and cognitive strategies, it is useful
to enquire how patients are currently trying to put the event behind them, what they think is
the best way of coping with the trauma, what they avoid, how they deal with intrusions, what
they think will happen if they allow themselves to dwell on the trauma or get upset about it,
whether they ruminate and what their ruminations consist of.
A further aim of the assessment interview is to start to characterise the nature of the trauma
memory and the spontaneous intrusions. Key issues include the extent to which there are gaps
in memory, whether the sequence of events seems muddled or confused and the extent to
which the memory/intrusions have a `here and now' quality and strong sensory and motor
components. Some of this information only becomes fully clear when some form of reliving
has been initiated (see below).
3.2. Rationale for treatment
Usually, the rationale for treatment has three elements. First, it is explained that PTSD
symptoms (especially intrusions, numbing and hyperarousal) are a common initial reaction to
an abnormal event. This point is emphasised by reviewing the patient's symptoms in detail and
explaining how some of the most puzzling aspects of the symptoms (e.g. the `here and now'-
quality of memories or becoming emotional for no apparent reason) are hallmarks of the
condition. Second, that many of the ways the patient has so far used to deal with the trauma
memory may have been useful for coping with other, milder stressors in their life, but
paradoxically may be maintaining their symptoms in this instance. Third, treatment involves
fully processing the trauma and reversing their particular maintaining factors.
A key element of treatment will involve thinking about the trauma more and discussing it in
detail. Various analogies can help explain this point. The therapist may compare the trauma
memory to a cupboard in which many things have been thrown in quickly and in a
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345336
disorganised fashion, so it is impossible to fully close the door and things fall out at
unpredictable times. Organising the cupboard will mean looking at each of the things and
putting them into their place. Once this is done, the door can be closed and remains shut.
Another useful analogy is that of a jigsaw puzzle that has been scattered all over the ¯oor so
that one will unexpectedly stumble over some of its pieces. Only when all the pieces have been
looked at and put together, the puzzle can be ®led away. Linked to these points, the therapist
explains that the reexperiencing symptoms are isolated memory fragments that are triggered by
matching cues and that they are experienced as if things were happening in the `here and now'
because they are not integrated which other autobiographical information.
3.3. Thought suppression experiment
For many patients who attempt to deal with intrusions by pushing them out of their mind, a
thought suppression experiment can be a useful way of illustrating the problematic
consequences of this strategy. For example, the therapist might say to the patient ``It doesn't
matter what you think for the next few minutes as long as you don't think about one
particular thing. It is extremely important you don't think about that thing... The thing is a
¯uorescent green bunny rabbit eating my hair!''. Most patients ®nd they immediately get an
image of the rabbit and have diculty getting rid of it. Discussion then helps them see that an
increase in the frequency of target thoughts is a normal consequence of thought suppression.
This result can then be used to set up a homework assignment in which the patient is asked to
collect data to test the idea that thought suppression may be enhancing intrusions. The
experiment involves not trying to push the intrusions out of the mind, but instead just letting
them come and go, watching them as though they were a train passing through a station.
Often patients report that this simple experiment produces a decline in both the frequency of
intrusions and the belief that intrusions are a sign of impending insanity or loss of control. A
similar approach can be used for rumination.
3.4. Education
Education about police, ambulance and hospital procedures, medication and other matters
can help correct many other problematic appraisals. For example, a patient thought that his
body was permanently damaged by an accident despite negative medical investigations. His
evidence for this conclusion was the fact that his urine had a very dark yellow colour for a few
days after the accident. He was greatly relieved to learn that this had been the eect of the
medication he had received.
3.5. Reclaiming one's life
A corollary of the here and now sense of intrusions is that patients with persistent PTSD feel
that their life is stuck at the time of the trauma (see Section Frozen in time). They often give
up important activities or social contacts that used to give them a sense of meaning and well-
being prior to the trauma. To help contextualise the memory and give patients the feeling that
they are moving forward in their lives they are encouraged to `reclaim' their former selves by
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 337
reinstituting activities that have dropped out of their lives. Often quite minor changes (e.g.
buying a new pair of trainers and going jogging again) can help reduce the feeling of being
stuck in time. If long-term physical eects of the trauma prevent taking up the original
activity, similar but manageable activities are explored. When planning the reactivation of
activities, it is important to identify problematic beliefs that may prevent the patient from
complying. For example, a patient who had a second motorbike accident after agreeing with
his family that he would not ride again, avoided visiting them because he was concerned that
they would reject him. Socratic questioning helped him see that this would not be the case.
3.6. Reliving with cognitive restructuring
Some form of reliving of the traumatic event is involved in most cognitive behavioural
programmes for PTSD. Procedures that have been shown to be eective include reliving the
experience in the presence of the therapist and putting this experience into words (e.g. Foa &
Rothbaum, 1998) or writing a detailed account of the event (e.g. Resick & Schnicke, 1993).
From the point of view of the present model, reliving has several important functions. First, it
promotes the elaboration and contextualisation of the trauma memory (see also Foa & Riggs,
1993). Second, identifying and discussing hot spots during reliving is useful in identifying the
idiosyncratic appraisals of the trauma. Third, for those patients who believe that they will go
crazy, fall apart, lose control or die when thinking about the trauma in detail, imaginal reliving
in itself is a powerful behavioural experiment to test this interpretation (see also Foa & Riggs,
Following careful explanation of the rationale for reliving (see overfull cupboard and other
metaphors above), we have tended to follow the general style of reliving recommended by Foa
and Rothbaum (1998), with some variants. Patients are instructed to relive the trauma in their
mind's eye, making the image as realistic as possible and including their thoughts and feelings
as well as what was happening. At the same time, they are asked to verbally describe the
reliving and to do so in the present tense. To help patients to stay with the memory, the
therapist asks questions such as ``What do you see?'', ``How does that feel?'', ``Where do you
feel that?'', ``What is going through your mind?''. To help identify hot spots, patients rate their
distress levels at dierent points during the reliving. Initially, reliving usually involves the
whole event, starting just before the event and continuing until patients knew they were safe.
As therapy progresses, reliving focuses more exclusively on hot spots and other problematic
aspects of the memory.
After a reliving exercise, therapist and patient identify and discuss problematic thoughts and
beliefs that are associated with the key moments of the trauma, using the relevant cognitive
restructuring techniques. Once an alternative perspective has been identi®ed, eorts are made
to incorporate this information into the next reliving. This can be achieved by carefully
reviewing the alternative interpretation before restarting reliving and practising answering one's
own thoughts during the reliving. In some cases, special techniques may be required. For
example, a patient who was devastated by the sexual response she experienced during a
protracted rape by a stranger was helped to see that although involuntary, this response was
probably the main reason why she was not killed and so could return to her husband and their
normal life. She had diculty incorporating this information at the relevant point during
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345338
reliving as she tended to dissociate. To get around this problem, she recorded her reappraisal
on tape and played it back through headphones at the relevant moment.
As therapy progresses, the nature of the trauma memory often changes. The narrative tends
to become more coherent, sensory components (e.g. smells, tastes and vivid images) and motor
components (e.g. involuntary movements) tend to fade and the memory loses its here and now
quality and becomes more like a normal recollection. For some patients, these changes occur
simply as a function of repeated reliving with the relevant rationale. For others, considerable
additional cognitive restructuring is required. Integrating reliving and cognitive restructuring
can be a challenge but, in our experience, can substantially reduce the amount of reliving
required for recovery. When integrating the two procedures, it is important to strike a balance
between sensitively spotting and changing appraisals and ensuring enough reliving to fully
activate the emotional components of the memory. Reliving is emotionally draining and care
needs to be taken to ensure that restructuring is not conducted when the patient is too
exhausted to bene®t.
Patients who are particularly likely to require extensive verbal and imagery cognitive
restructuring are those who: (1) experience anger, guilt or shame as a predominant emotion, (2)
interpret their behaviour or emotions during the event as showing something negative about
themselves (e.g. perpetrators of crime (Foa & Meadows, 1997), rape victims who experience
mental defeat (Ehlers et al., 1998a)) or (3) experienced violence over a prolonged period of
time. The latter group sometimes cannot help but assume the perpetrators' negative views
about them to some extent, viewing themselves as criminals or deserving maltreatment
(Saporta & van der Kolk, 1992; Reemtsma, 1997; Ehlers et al., in press). For some individuals
in these categories, extensive cognitive restructuring may be required before imaginal reliving
can be bene®cial.
At this stage, it is unclear why reliving works. However, there are several ways in which it is
likely to facilitate elaboration of the trauma memory. First, it links previously unconnected
parts of the traumatic experience, thus giving them a context. This will reduce the probability
that isolated parts of the memory are triggered. A woman whose young daughter died in a
house ®re while she was out, had frequent intrusions of seeing the curtains burning when she
approached the house. At the time, she had thought that the daughter was burning alive and
was in tremendous pain. However, the daughter had actually been upstairs at the time and the
®re had not reached her (she had died from the fumes), a fact that the patient took great
comfort in. She had for years avoided thinking about the event and had never connected the
fact that the daughter was upstairs with the image of the curtains burning. When she
connected the image of the curtains burning with the image of the daughter in the upstairs
bedroom in imaginary, her intrusions of the burning curtains ceased.
Second, reliving (as well as in vivo inspection of the site of the trauma) facilitates the
retrieval of elements of the trauma memory that are dicult for the patient to access
otherwise. In some cases accessing the previously unretrieved information leads to immediate
changes in the problematic appraisals. For example, a patient was extremely angry with the
paramedics who rescued her from her car after an accident because they did not answer her
question of whether she was going to be paralysed. At the time, she had interpreted this as
meaning they did not regard her as a human being. During reliving, she realised that the
paramedics were probably concerned about upsetting her because she had been very agitated
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 339
before and they had only just managed to calm her down as to try to get her out. Once she
had accessed this information and changed her interpretation, her distress ratings during
reliving changed dramatically and her intrusions of being trapped in the car ceased.
Third, patients may link information they received after the trauma to correct their
impression and thoughts during the trauma so that the event poses less current threat to the
self. For example, a bus driver who had run over an elderly lady and felt very guilty became
increasingly aware during reliving that the lady had intended to commit suicide by stepping out
in front of the bus and his intrusions of seeing the lady look at him shortly before the impact
Fourth, reliving facilitates the discrimination between the `then' and `now', i.e.
discrimination of how the stimulus con®guration during the traumatic event diered from
those during other safe events (see also Foa & Rothbaum, 1998). Thus, with the elaboration a
closer match between the original traumatic situation and current situations will be required
for a memory to be triggered.
Fifth, the verbalisation of visual and other sensory cues may also make it more dicult to
retrieve the original sensory impressions from memory
3.7. In vivo exposure
In vivo exposure to avoided reminders of the trauma (e.g. the site, similar situations,
activities, feelings, smells and sounds) is a powerful way of helping patients to emotionally
accept that the traumatic event is in the past. When revisiting the site of the event, discussion
of similarities and dierences between what the scene looked like during the trauma and what
it looks like now helps the patient in establishing a time perspective and helps in discriminating
the harmless stimuli that happened to coincide with the trauma from the dangerous stimuli
encountered during the traumatic event. Revisiting the site can also provide new information
which helps correct problematic appraisals (e.g. seeing the road layout and discovering that
one could not have prevented an accident).
Overgeneralisation of danger (e.g. never going out at night or drinking alcohol after being
raped on a night out) can be eectively challenged by setting up exposure to avoided activities
as a behavioural experiment (see also Clark, 1999). Patients are asked to specify what is the
worst they think could happen and how likely it seems before entering the avoided situation/
engaging in the avoided activity. In order to maximise the possibility of discon®rmation,
patients are also encouraged to drop any relevant safety behaviours. For example, a driver
who repeatedly looked in the mirror, turned o the radio (to facilitate hyperattention to the
road) and ®rmly gripped the steering wheel in order to prevent future accidents would be
encouraged to drop all of these behaviours and return to a pre-accident driving pattern.
Appraisals of trauma sequelae can also be challenged by setting up in vivo exposure as a
behavioural experiment. For example, a patient who found herself becoming emotional and
irritable for no apparent reason after a severe road trac accident in which she had been
Experiments have shown that giving verbal descriptions of pictures decreases the ability to identify the pictures
(verbal overshadowing: Schooler & Engstler-Schooler, 1990).
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345340
trapped in her car, interpreted her reactions as meaning that she would go crazy if she put
herself under stress again and, as a consequence, would become like her sister who suered
from schizophrenia. A behavioural experiment was set up that involved thinking about the
accident in an emotional way while in a car wash (a strong reminder of being trapped in her
car). Before the behavioural experiment, the patient believed 100% that thinking about the
accident in the car wash would make her go crazy. Her belief changed dramatically when she
found that the experience was tolerable and that there were no signs of her going crazy.
3.8. Identifying triggers of intrusive memories and emotions
The model suggests that one way by which the elaboration of the trauma memory reduces
the probability of reexperiencing symptoms is by promoting a better discrimination between
those stimuli that occurred around the time of the trauma and those encountered currently.
This process can be enhanced by direct interventions aiming at better discrimination. First,
patients may bene®t from training in spotting triggers of intrusive memories or negative aect
and physical sensations related to the trauma. This requires careful monitoring of occasions
when intrusions occur and information about the likely nature of the triggers (e.g. physical
cues that were temporally associated with the trauma, but may not have a strong semantic
relationship to the trauma: lights, smells, touch, movement, etc). Once the patient has identi®ed
triggers, detailed discussion of the similarities and dierences between the present and past
(trauma-related) context of the triggers can be used to facilitate stimulus discrimination. For
example, a rape victim reported that she had been feeling very uneasy when having sex with
her husband, even though she was not recalling the rape at the time. Therapist and patient
discussed in detail the way the rapist had behaved and her husband's behaviour during sex. It
emerged that their were quite a few sensory similarities, e.g. the way both men touched certain
parts of her body, both events taking place in the dark and being accompanied by talking.
Next, they discussed the dierences, with particular emphasis on the men's intentions and their
attitude to her. In this way, the patient was able to see that the similar sensory cues had very
dierent meanings in the two contexts. To help further promote discrimination between the
two events she was instructed to pay particular attention to things that were dissimilar from
the rape when having sex with her husband and to change some of the stimulus conditions
(e.g. leaving light on) to facilitate discrimination.
3.9. Imagery techniques
Imagery techniques are also useful in elaborating and changing the meaning of the trauma
memory. For example, a person whose friend was blown up was unable to mentally say
goodbye to the friend until he visualised him dead but whole again. A man who was hit head-
on by another car felt guilty because he believed that the other driver must have been in
tremendous protracted agony in the awareness of impending death, outweighing any distress
that he had experienced. When he visualised the accident from the other driver's perspective he
became aware that she must have only seen his car very shortly before the impact and must
have died immediately. Imagery also allows patients to explore the possible consequences of
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 341
actions that were not taken and to incorporate a spiritual viewpoint (Layden & Hackmann, in
4. Summary and empirical support
It is suggested that PTSD becomes persistent when individuals process the trauma in a
way which produces a sense of serious, current threat. The sense of threat arises as a
consequence of: (1) excessively negative appraisals of the trauma and/or its sequelae and
(2) a disturbance of autobiographical memory characterised by poor elaboration and
contextualisation, strong associative memory and strong perceptual priming. Change in the
negative appraisals and the trauma memory are prevented by a series of problematic
behavioural and cognitive strategies.
The proposed model is consistent with the main clinical features of PTSD, helps explain
several apparently puzzling phenomena (the `here and now' quality of the memory and
intrusions; `aect without recollection', delayed onset PTSD, problems in intentional recall and
easily triggered reexperiencing) and provides a framework for treatment by identifying three
key targets for change.
Many propositions in the model remain to be tested. However, it is encouraging to note that
recent studies have provided support for several central features. In particular, (1) negative
appraisals of the trauma (Dunmore et al., 1997, 1998, in press; Foa et al., in press), negative
interpretations of initial PTSD symptoms (Ehlers & Steil, 1995; Dunmore et al., 1997, 1998,
1999; Ehlers, Mayou & Bryant, 1998; Clohessy & Ehlers, in press; Steil & Ehlers, in press) and
negative interpretations of other people's posttrauma responses (Dunmore et al., 1997, 1998,
1999; Ehlers et al., in press) been have shown to predict PTSD persistence; (2) Foa and
colleagues found that degree of improvement during cognitive-behavioural treatment is related
to the extent to which the trauma narrative becomes more organized and coherent (Amir et al.,
1998; Foa et al., 1995); (3) Murray, Ehlers and Mayou (submitted) found that memory
fragmentation predicted PTSD persistence; (4) analogue experiments demonstrated enhanced
perceptual priming for stimuli that occur in a traumatic context (Ehlers, Michael & Chen, in
preparation) and (5) several strategies highlighted in the model (thought suppression,
rumination, safety behaviours and avoidance) have been shown to predict persistence
(Dunmore et al., 1998, 1999; Ehlers et al., 1998a, 1998b; Clohessy & Ehlers, in press; Steil &
Ehlers, in press; Murray et al., submitted for publication). It is hoped that future studies will
further investigate the model and its implications for treatment.
A.E. and D.M.C. are Wellcome Principal Research Fellows. We are grateful to Emma
Dunmore, Melanie Fennell, Ann Hackmann, Freda McManus and Regina Steil for their
collaboration, ideas and insightful clinical observations that are given as examples in this
paper. They made many important contributions to the conceptualization of PTSD treatment
outlined here. We thank Edna B. Foa for many inspiring discussions and her collaboration.
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345342
Many thanks to Martin Conway and Chris Brewin for their suggestions. We thank Ann
Hackmann, Freda McManus, Melanie Fennell and Warren Mansell for their helpful comments
on earlier drafts of this manuscript.
Amir, N., Staord, J., Freshman, M. S., & Foa, E. B. (1998). Relationship between trauma narratives and trauma
pathology. Journal of Traumatic Stress,11, 385±392.
Baddeley, A. (1997). Human memory. Theory and practice (revised ed.). Hove, UK: Psychology Press.
Bechara, A., Trandel, D., Damasio, H., Adelphs, R., Rockland, C., & Damasio, A. R. (1995). Double dissociation
of conditioning and declarative knowledge relative to the amygdalae and hippocampus in humans. Science,269,
Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: International Universities Press.
Brewin, C. R., Dalgleish, T., & Joseph, S. (1996). A dual representation theory of posttraumatic stress disorder.
Psychological Review,103, 670±686.
Charney, D., Deutch, A. Y., Krystal, J. H., Southwick, S. M., & Davis, M. (1993). Psychobiological mechanisms of
posttraumatic stress disorder. Archives of General Psychiatry,50, 294±305.
Clark, D. M. (1999). Anxiety disorders: why they persist and how to treat them. Behaviour Research and Therapy,
37, S5±S27.
Clohessy, S. & Ehlers, A. (in press). PTSD symptoms, response to intrusive memories and coping in ambulance
service workers. British Journal of Clinical Psychology.
Conway, M. A. (1997a). Introduction: what are memories? In M. A. Conway, Recovered memories and false
memories (pp. 1±22). Oxford, UK: Oxford University Press.
Conway, M. A. (1997b). Past and present: recovered memories and false memories. In M. A. Conway, Recovered
memories and false memories (pp. 150±191). Oxford, UK: Oxford University Press.
Conway, M. A., & Pleydell-Pearce, C. W. (1997). On the construction of autobiographical memories: the self-
monitoring system and its neuroanatomical basis.
Davey, G. C. L. (1989). UCS revaluation and conditioning models of acquired fears. Behaviour Research and
Therapy,27, 521±528.
Dunmore, E., Clark, D. M., & Ehlers, A. (1997). Cognitive factors in persistent versus recovered posttraumatic
stress disorder after physical or sexual assault: a pilot study. Behavioural and Cognitive Psychotherapy,25, 147±
Dunmore, E., Clark, D. M. & Ehlers, A. (1998). The role of cognitive factors in posttraumatic stress disorder
following physical or sexual assault: ®ndings from retrospective and prospective investigations. Paper presented
at Annual Conference of British Association of Behavioural and Cognitive Therapies. Durham, UK, July 9±11.
Dunmore, E., Clark, D. M. & Ehlers, A. (1999). Cognitive factors involved in the onset and maintenance of PTSD.
Behaviour Research and Therapy, 37, 809±829.
Ehlers, A., Clark, D. M., Dunmore, E., Jaycox, L., Meadows, E., & Foa, E. B. (1998a). Predicting response to
exposure treatment in PTSD: the role of mental defeat and alienation. Journal of Traumatic Stress,11, 457±471.
Ehlers, A., Hackmann, A., Steil, R., Clohessy, S. & Wenninger, K. (1999). On the nature of posttraumatic
intrusions. Manuscript in preparation.
Ehlers, A., Maercker, A. & Boos, A. (in press). PTSD following political imprisonment: The role of mental defeat,
alienation and perceived permanent change. Journal of Abnormal Psychology.
Ehlers, A., Mayou, R. A., & Bryant, B. (1998b). Psychological predictors of chronic posttraumatic stress disorder
after motor vehicle accidents. Journal of Abnormal Psychology,107, 508±519.
Ehlers, A., Michael, T. & Chen, Y. P. (in preparation). Enhanced perceptual priming for stimuli that occur in a
traumatic context.
Ehlers, A., & Steil, R. (1995). Maintenance of intrusive memories in posttraumatic stress disorder: a cognitive
approach. Behavioural and Cognitive Psychotherapy,23, 217±249.
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 343
Foa, E. B., Ehlers, A., Clark, D. M., Tolin, D. F. & Orsillo, S. M. (in press). The posttraumatic cognitions
inventory (PTCI): development and validation. Psychological Assessment.
Foa, E. B., & Hearst-Ikeda, D. (1996). Emotional dissociation in response to trauma: an information processing
approach. In L. K. Michelson, & W. J. Ray, Handbook of dissociation: theoretical, empirical and research
perspectives (pp. 207±224). New York: Plenum Press.
Foa, E. B., & Kozak, M. J. (1986). Emotional processing of fear: exposure to corrective information. Psychological
Bulletin,99, 20±35.
Foa, E. B., & Meadows, E. A. (1997). Psychosocial treatments for posttraumatic stress disorder: a critical review.
Annual Review of Psychology,48, 449±480.
Foa, E. B., Molnar, C., & Cashman, L. (1995). Change in rape narratives during exposure therapy for
posttraumatic stress disorder. Journal of Traumatic Stress,8, 675±690.
Foa, E. B., & Riggs, D. S. (1993). Post-traumatic stress disorder in rape victims. In J. Oldham, M. B. Riba, & A.
Tasman, Annual review of psychiatry,Vol. 12 (pp. 273±303). Washington, DC: American Psychiatric Association.
Foa, E. B., & Rothbaum, B. O. (1998). Treating the trauma of rape. Cognitive-behavior therapy for PTSD. New
York: Guilford.
Foa, E. B., Steketee, G., & Rothbaum, B. O. (1989). Behavioural/cognitive conceptualisations of post-traumatic
stress disorder. Behaviour Therapy,20, 155±176.
Herman, J. L. (1992). Complex PTSD: a syndrome in survivors of prolonged and repeated trauma. Journal of
Traumatic Stress,5, 377±391.
Horowitz, M. J. (1997). Stress response syndromes. PTSD, grief and adjustment disorders. Northvale, NJ: Jason
Jano-Bulman, R. (1992). Shattered assumptions: toward a new psychology of trauma. New York: The Free Press.
Joseph, S., Williams, R., & Yule, W. (1997). Understanding posttraumatic stress. A psychosocial perspective on PTSD
and treatment. Chicester, UK: Wiley.
Jones, J. C., & Barlow, D. H. (1990). The etiology of posttraumatic stress disorder. Clinical Psychology Review,10,
Keane, T. M., Zimering, R. T., & Caddell, J. M. (1985). A behavioral formulation of posttraumatic stress disorder.
The Behavior Therapist,8, 9±12.
Kessler, R. C., Sonnega, A., Bromet, E., Hughes, M., & Nelson, C. B. (1995). Posttraumatic stress disorder in the
National Comorbidity Survey. Archives of General Psychiatry,52, 1048±1060.
Koss, M. P., Figueredo, A. J., Bell, I., Tharan, M., & Tromp, S. (1996). Traumatic memory characteristics: a cross-
validated mediational mode of response to rape among employed women. Journal of Abnormal Psychology,105,
Krystal, J. H., Bennett, A. L., Bremner, J. D., Southwick, S. M., & Charney, D. S. (1995). Toward a cognitive
neuroscience of dissociation and altered memory functions in posttraumatic stress disorder. In M. J. Friedman,
D. S. Charney, & A. Y. Deutch, Neurobiological and clinical consequences of stress: from normal adaptation to
PTSD (pp. 239±269). Philadelphia, PA: Lippincott-Raven Publishers.
Kuyken, W., & Brewin, C. R. (1995). Autobiographical memory functioning in depression and reports of early
abuse. Journal of Abnormal Psychology,104, 585±591.
Layden, M. & Hackmann, A. (in preparation). Imagery and cognitive therapy. Washington, DC: American
Psychological Association.
LeDoux, J. E. (1992). Emotion as memory: anatomical systems underlying indelible memory traces. In S. A.
Christianson, Handbook of emotion and memory (pp. 269±288). Hillsdale, NJ: Erlbaum.
Markowitsch, H. J. (1995). Which brain regions are critically involved on the retrieval of old episodic memory?
Brain Research Reviews,21, 117±127.
Markowitsch, H. J. (1996). Organic and psychogenic retrograde amnesia: two sides of the same coin? Neurocase,2,
Markowitsch, H. J., Kessler, J., Van der Ven, C., Weber-Luxemburger, G., Albers, M., & Heiss, W. D. (1998).
Psychic trauma causing grossly reduced brain metabolism and cognitive deterioration. Neuropsychologia,36, 77±
McNally, R. J., Lasko, N. B., Macklin, M. L., & Pitman, R. K. (1995). Autobiographical memory disturbance in
combat-related posttraumatic stress disorder. Behaviour Research and Therapy,33, 619±630.
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345344
McNally, R. J., & Shin, L. M. (1995). Association of intelligence with severity of posttraumatic stress disorder
symptoms in Vietnam combat veterans. American Journal of Psychiatry,152, 936±938.
Meichenbaum, D. (1997). Treating posttraumatic stress disorder. A handbook and practice manual for therapy.
Chichester, UK: Wiley.
Murray, J., Ehlers, A. & Mayou, R. M. (submitted for publication). Two prospective studies of PTSD following
motor verhicle accidents.
Newcomer, J. W., Selke, G., Melson, A. K., Hershey, T., Craft, S., Richards, K., & Alderson, A. L. (1999).
Decreased memory performance in healthy humans induced by stress-level cortisol treatment. Archives of General
Psychiatry,56, 527±533.
Pennebaker, J. W. (1989). Confession, inhibition and disease. Advances in Experimental Social Psychology,22, 211±
Reemtsma, J. P. (1997). Im Keller (In the cellar). Hamburg, Germany: Hamburger Edition.
Rescorla, R. A. (1988). Pavlovian conditioning: it's not what you think it is. American Psychologist,43, 151±160.
Resick, P. A., & Schnicke, M. K. (1993). Cognitive processing therapy for rape victims. Newbury Park, CA: Sage.
Roediger, H. L. (1990). Implicit memory: retention without remembering. American Psychologist,45, 1043±1056.
Rothbaum, B. O., Foa, E. B., Riggs, D. S., Murdock, T. B., & Walsh, W. (1992). A prospective examination of
posttraumatic stress disorder in rape victims. Journal of Traumatic Stress,5, 455±475.
Salkovskis, P. M. (1996). The cognitive approach to anxiety: threat beliefs, safety-seeking behaviour and the special
case of health anxiety and obsessions. In P. M. Salkovskis, Frontiers of cognitive therapy (pp. 48±74). New York:
Saporta, J. A., & van der Kolk, B. A. (1992). Psychobiological consequences of severe trauma. In M. Basoglu,
Torture and its consequences (pp. 151±181). Cambridge, UK: Cambridge University Press.
Schacter, D. L., Norman, K. A., & Koutstaal, W. (1997). The recovered memories debate: A cognitive neuroscience
perspective. In M. A. Conway, Recovered memories and false memories (pp. 63±99). Oxford, UK: Oxford
University Press.
Schooler, J. W., & Engstler-Schooler, T. Y. (1990). Verbal overshadowing of visual memories: some things are
better left unsaid. Cognitive Psychology,22, 36±71.
Siegel, D. J. (1995). Memory, trauma and psychotherapy: a cognitive science view. Journal of Psychotherapy
Practice and Research,4, 93±122.
Spiegel, D. (1991). Dissociation and trauma. In A. Tasman, & S. M. Gold®nger, Annual review of psychiatry,Vol.
10 (pp. 261±275). Washington, DC: American Psychiatric Press.
Steil, R. & Ehlers, A. (in press). Dysfunctional meaning of posttraumatic intrusions in chronic PTSD. Behaviour
Research and Therapy.
van der Kolk, B. A., & Fisler, R. (1995). Dissociation and the fragmentary nature of traumatic memories: overview
and exploratory study. Journal of Traumatic Stress,8, 505±525.
van der Kolk, B. A., & van der Hart, O. (1991). The intrusive past: the ¯exibility of memory and the engraving of
trauma. American Imago,48, 425±454.
Wheeler, A. M., Stuss, D. T., & Tulving, E. (1997). Toward a theory of episodic memory: the frontal lobes and
autonoetic consciousness. Psychological Bulletin,121, 331±354.
Wegner, D. M. (1989). White bears and other unwanted thoughts: suppression, obsession and the psychology of mental
control. New York: Viking.
A. Ehlers, D.M. Clark / Behaviour Research and Therapy 38 (2000) 319±345 345
... Current prominent theories of PTSD (Brewin et al., 1996a;Brewin & Holmes, 2003;Dalgleish, 2004;Ehlers & Clark, 2000;Foa & Rothbaum, 2001) share the assumption that traumatic experiences are encoded in more than one memory representation. Evaluating these theories, Dalgleish (2004) as well as Brewin (2011) conclude that it is necessary to distinguish between three different types of memory representation: the sensation-based memory of the event, that codes the perceptual elements of the experience in the form of associative networks; the contextualized episodic memory, that provides a temporal and spatial context of the perceptual memory; and the schema based self-concept, that encodes knowledge and beliefs about the self in relation to the social world. ...
... Other theories refer to this association as the fear structure (Foa and Kozak 1986). The sensation-based memory is generally modeled as an excessive associative network that encodes abnormally strong connections between the stimulus and response elements that had been present during the traumatic event (Ehlers & Clark, 2000;Foa & Rothbaum, 2001). This associative network includes the sensory (visual, auditory, and other senses) as well as physiological and behavioral responses and can be activated by the perception of a sensory stimulus that resembles an item of the associative network. ...
... Within the PAST framework, the psychopathology of trauma-related disorders does not only result from the uncontrolled activation of the threat structure, but also from dysfunctional attempts to cope with the trauma memory. One of the most common dysfunctional coping strategies is trying to avoid confrontation with reminders of the traumatic event (Ehlers & Clark, 2000). This includes overt avoidance strategies related to salient triggers of traumatic memory, for example sexual activity for victims of sexual abuse, but also implicit avoidance of potentially threatening stimuli (Fleurkens et al., 2014). ...
Full-text available
Current theories of psychological trauma assume that posttraumatic symptoms originate from stress reactions caused by extremely adverse life experiences. Since the diagnosis of PTSD is restricted to events that involve threats to the physical or sexual integrity of a person, such as accidents and physical and sexual violence, these theories are not well suited to explain the psychopathological consequences of severe violations of one's social integrity, such as emotional abuse and bullying. However, it is evident that social threats contribute to a broad range of mental disorders and increase symptom severity in patients with posttraumatic stress disorder. The aim of the Physical and Social Trauma (PAST) framework is to extend current memory theories of psychological trauma to incorporate threats to a person's social integrity. Within this perspective, the harmful effects of events that involve social threats result from violations of core social motives such as the need for status and belonging that bring about intense affective reactions, including despair and defeat. Within associative threat structures, these emotions are tied to the stimulus characteristics of the experiences and can be re-activated in social situations. The resulting psychopathology transcends PTSD criteria and other current classifications and suggests a transdiagnostic perspective of psychological trauma. Implications for treatment and further directions for research are discussed.
... RTM therapists undertook precourse reading, 40 hours over four days of face-to-face classroom teaching and four hours of symptom assessment and therapy delivery on two trauma patients which were observed and assessed for delity by the RTM trainer/supervisor and an external assessor from the US research team. TFCBT therapists undertook 24 hours of face-to-face classroom teaching aligned to Ehlers and Clark's cognitive processing model [21,22] incorporating re ective exercises alongside practical clinical case examples illustrating key intervention strategies. Their competency was assessed as the training progressed. ...
... TFCBT was delivered over up to 18 x 60-90 minute weekly sessions [22]. The cognitive processing model of PTSD developed by Ehlers and Clark (2000) was purposely used in this study because it displays the largest treatment effect sizes and signi cant symptom improvement and is widely delivered through IAPT services [21,22,26]. TFCBT is focused on identifying the relevant appraisals, memory characteristics and triggers, and behavioural and cognitive strategies that maintain PTSD symptoms. ...
... TFCBT was delivered over up to 18 x 60-90 minute weekly sessions [22]. The cognitive processing model of PTSD developed by Ehlers and Clark (2000) was purposely used in this study because it displays the largest treatment effect sizes and signi cant symptom improvement and is widely delivered through IAPT services [21,22,26]. TFCBT is focused on identifying the relevant appraisals, memory characteristics and triggers, and behavioural and cognitive strategies that maintain PTSD symptoms. ...
Full-text available
Background: Posttraumatic stress disorder (PTSD) occurs more commonly in military veterans than the general population. Whilst current therapies are effective, up to half of veterans commencing treatment don’t complete it. Reconsolidation of Traumatic Memories (RTM) protocol is a novel, easy to train, talking therapy with promising US findings. We examine the feasibility of undertaking an efficacy trial of RTM in UK veterans. Methods: A parallel group, single-centre randomised controlled feasibility trial with a post-completion qualitative interview study. Sixty UK military veterans were randomised 2:1 to RTM (n=35) or Trauma Focussed CBT (n=25). We aimed to determine the rate of recruitment and retention, understand reasons for attrition, determine data quality and size of efficacy signal. We explored veterans’ perceptions of experiences of joining the trial, the research procedures and therapy, and design improvements for future veteran studies. UK military veterans with a diagnosis of PTSD or complex PTSD, and clinically significant symptoms, were recruited between January 2020 and June 2021. Primary outcome was feasibility using pre-determined progression criteria alongside PTSD symptoms, with depression, recovery and rehabilitation as secondary outcomes. Data were collected at baseline, 6, 12 and 20 weeks. Interviews (n=15) were conducted after 20 weeks. Both therapies were delivered by trained charity sector provider therapists. Results: Participants’ mean age was 53yrs, the mean baseline PTSD symptoms score assessed by PCL-5 was 57. Fifty had complex PTSD and 39 had experienced ≥4 traumas. Data were analysed at 20 weeks for feasibility outcomes (n=60) and mental health outcomes (n=45). Seven of eight progression criteria were met. The RTM group experienced a mean 18-point reduction on the PCL-5. TFCBT group participants experienced a mean reduction of 8 points. 48% of the RTM group no longer met diagnostic criteria for PTSD compared to 16% in the TFCBT group. All veterans reported largely positive experiences of the therapy and research procedures and ways to improve them. Conclusion: RTM therapy remains a promising psychological intervention for the treatment of PTSD, including complex PTSD, in military veterans. With specific strengthening, the research protocol is fit for purpose in delivering an efficacy trial. Trial registration: ISRCTN registration no 10314773 on 01.10.2019
... As this example shows, in many cases, the trigger stimuli do not even have a strong conceptual relationship to the traumatic event, but are simply cues that were temporally associated with it, especially those present shortly before or during the traumatic event, such as smells and objects, specific phrases said in a certain tone of voice, etc. (Ehlers and Clark, 2000). ...
... Because of this, recall of traumatic events may be poorly organized, fragmented, and not integrated into the autobiographical memory base. Sensory imagery can continue to be re-experienced even if other information that contradicts the imagery, suggesting it is false, has already been encoded in memory (Ehlers & Clark, 2000). Furthermore, traumatic memories are less likely than other memories to be quickly updated in response to information pertaining to the event that is 2 The idea that memories of traumatic experiences are fragmented and poorly organized continues to be the object of debate among psychologists (see Brewin 2007;Brewin, 2016;Rubin et al, 2016;and more recently, McNally, 2021;McNally, 2022). ...
Fear generalisation is a psychological phenomenon that occurs when fear that is elicited in response to a frightening stimulus spreads to similar or related stimuli. The practical harms of fear generalisation are well-documented, but little or no attention has been given so far to its epistemic harms. This paper fills this gap in the literature. It shows how the psychological phenomenon substantially curbs the epistemic agency of those who experience it, limiting their ability to respond to evidence, and substantially limiting their epistemic horizons. It is argued that when these epistemic harms are caused by wrongful actions and decisions of individuals or institutions, because the fear is elicited in response to a traumatic experience inflicted by them, the harms should be considered epistemic wrongs. The epistemic wrongs are closely akin to agential epistemic injustice, a variety of distributive epistemic injustice, and sometimes also involve epistemic exclusion. The paper thereby identifies a previously underexplored psychological mechanism that can be a vehicle through which both individuals and institutions can epistemically wrong others. The argument has implications for how both epistemic wrongs and epistemic injustice should be conceived, suggesting that both can occur without primarily being caused by epistemic flaws or errors, or a bad epistemic character. Finally, it highlights the advantage of taking a victim-centred approach to understanding epistemic harm.
... They may expect people in the present to respond to them in the same way people might have done during earlier experiences of excessive criticism, bullying, racism or victimisation. Then vs Now discrimination training is a technique originally developed to treat trauma triggers in post-traumatic stress disorder (Ehlers and Clark, 2000) that we have found helpful with many patients with SAD over the years. It can help patients to process current social situations without being influenced by feelings or memories from past trauma. ...
Full-text available
Patients with social anxiety disorder (SAD) have a range of negative thoughts and beliefs about how they think they come across to others. These include specific fears about doing or saying something that will be judged negatively (e.g. ‘I’ll babble’, ‘I’ll have nothing to say’, ‘I’ll blush’, ‘I’ll sweat’, ‘I’ll shake’, etc.) and more persistent negative self-evaluative beliefs such as ‘I am unlikeable’, ‘I am foolish’, ‘I am inadequate’, ‘I am inferior’, ‘I am weird/different’ and ‘I am boring’. Some therapists may take the presence of such persistent negative self-evaluations as being a separate problem of ‘low self-esteem’, rather than seeing them as a core feature of SAD. This may lead to a delay in addressing the persistent negative self-evaluations until the last stages of treatment, as might be typically done in cognitive therapy for depression. It might also prompt therapist drift from the core interventions of NICE recommended cognitive therapy for social anxiety disorder (CT-SAD). Therapists may be tempted to devote considerable time to interventions for ‘low self-esteem’. Our experience from almost 30 years of treating SAD within the framework of the Clark and Wells (1995) model is that when these digressions are at the cost of core CT-SAD techniques, they have limited value. This article clarifies the role of persistent negative self-evaluations in SAD and shows how these beliefs can be more helpfully addressed from the start, and throughout the course of CT-SAD, using a range of experiential techniques. Key learning aims (1) To recognise persistent negative self-evaluations as a key feature of SAD. (2) To understand that persistent negative self-evaluations are central in the Clark and Wells (1995) cognitive model and how to formulate these as part of SAD. (3) To be able to use all the experiential interventions in cognitive therapy for SAD to address these beliefs.
... Interestingly, the other PCS subscales (rumination and magnifcation) were not signifcant mediators in any of the analyses. Te tendency to engage in catastrophic thinking has been long thought to serve as a precursor to PTSD symptoms following the experience of stressful events [31], and individuals with chronic pain and comorbid PTSD are more likely to experience less control over pain, heightened emotional responses to pain, and higher levels of catastrophizing compared to individuals with chronic pain alone [32]. Te fndings are consistent with previous research suggesting pain catastrophizing serves as a potentially important cognitive variable to consider when conceptualizing the relationship between chronic pain and PTSD symptoms [13,14,33]. ...
Full-text available
Background. Comorbid chronic pain and post-traumatic stress disorder (PTSD) complicate the treatment of both conditions. Previous research has identified pain catastrophizing as a potentially important variable contributing to the relationship between chronic pain and PTSD. However, little is known regarding how the different dimensions of pain catastrophizing—rumination, magnification, and helplessness—uniquely contribute to the relationship between PTSD symptomatology and measures of pain outcome. Methods. 491 treatment seeking participants were admitted to a three-week interdisciplinary pain rehabilitation program between July 2016 and March 2020. The patients completed measures of pain severity, pain interference, pain catastrophizing, depressive symptoms, quality of life (QOL), and PTSD symptoms at pretreatment. Results. Parallel mediation analyses were conducted to evaluate the mediating effect of the Pain Catastrophizing Scale subscales on the relationship between PTSD symptomatology and pain-relevant variables. The helplessness subscale accounted for significant unique variance in the relationship between PTSD symptomatology and pain severity (b = 0.010, SE = 0.002, 95% CI: 0.006, 0.014), pain interference (b = 0.004, SE = 0.002, 95% CI: 0.001, 0.008), and mental health QOL (b = −0.117, SE = 0.031, 95% CI: −0.179, −0.059), while the rumination and magnification subscales had no significant influence. Conclusions. Pain catastrophizing is a multifaceted construct. These results suggest that the helplessness dimension of pain catastrophizing may be the primary target when treating patients with comorbid chronic pain and PTSD symptoms. This study represents the first to evaluate the influence of the individual dimensions of pain catastrophizing on the relationship between PTSD symptomatology and chronic pain outcome.
... Current uses of CBT for the treatment of grief are limited to the use of post-traumatic stress disorder (PTSD) models of CBT (Ehlers & Clark, 2000) for the treatment of traumatic grief -which is a form of complicated grief resulting from death by traumatic or unexpected means (Howarth, 2011), the use of complicated grief therapy (CGT; Malkinson, 2007Malkinson, , 2010Malkinson et al., 2006) which has a foundation in both CBT for PTSD and interpersonal psychotherapy (IPT) for depression (Wetherell, 2012), and CBT-based self-help approaches such as the Overcoming Grief book by Sue Morris (2018). ...
... According to the cognitive model of PTSD (Ehlers & Clark, 2000), cognitive processes relating to the memory of the traumatic experience is central to the development and maintenance of chronic PTSD. It is also consistent with the findings where aspects of emotion regulation, in particular thought suppression and experiential avoidance (i.e., cognitive avoidance), demonstrated the strongest association with PTSD (Seligowski, Lee, (Trickey et al., 2012). ...
Background Youth receiving medical care for injury are at risk of PTSD. Therefore, accurate prediction of chronic PTSD at an early stage is needed. Machine learning (ML) offers a promising approach to precise prediction and interpretation. Aims The study proposes a clinically useful predictive model for PTSD 6-12 months after injury, analyzing the relationship among predictors, and between predictors and outcomes. Methods A ML approach was utilized to train models based on 1,167 children and adolescents of nine perspective studies. Demographics, trauma characteristics and acute traumatic stress (ASD) symptoms were used as initial predictors. PTSD diagnosis at six months was derived using DSM-IV PTSD diagnostic criteria. Models were validated on external datasets. Shapley value and partial dependency plot (PDP) were applied to interpret the final model. Results A random forest model with 13 predictors (age, ethnicity, trauma type, intrusive memories, nightmares, reliving, distress, dissociation, cognitive avoidance, sleep, irritability, hypervigilance and startle) yielded F-scores of.973,.902 and.961 with training and two external datasets. Shapley values were calculated for individual and grouped predictors. A cumulative effect for intrusion symptoms was observed. PDP showed a non-linear relationship between age and PTSD, and between ASD symptom severity and PTSD. A 43% difference in the risk between non-minority and minority ethnic groups was detected. Conclusions A ML model demonstrated excellent classification performance and good potential for clinical utility, using a few easily obtainable variables. Model interpretation gave a comprehensive quantitative analysis on the operations among predictors, in particular ASD symptoms.
Full-text available
Dysregulated consolidation of emotional memories is a core feature of posttraumatic stress disorder (PTSD). Brain Derived Neurotrophic Factor (BDNF) influences synaptic plasticity and emotional memory consolidation. The BDNF Val66Met polymorphism has been associated with PTSD risk and memory deficits respectively, although findings have been inconsistent, potentially due to a failure to control for important confounds such as sex, ethnicity, and the timing (child/adult) and extent of previous trauma experiences. Furthermore, very little research has examined the impact of BDNF genotypes on emotional memory in PTSD populations. This study investigated the interaction effects of the BDNF Val66Met genotype and PTSD symptomatology in an emotional recognition memory task in 234 participants. Key findings revealed impaired negative recognition memory in PTSD (n=44) compared to Control (n=85) and trauma exposed (n=105) groups and in participants with the Val/Met compared to the Val/Val genotype. There was a group x genotype interaction driven by the lack of impairment in the TE group with the Val/Met genotype. Results suggest that people previously exposed to trauma who do not develop PTSD may be protected from the BDNF Met effect, however more research is needed to replicate findings and to explore the relevant epigenetic and neural processes involved.
This chapter describes a cognitive model of the autobiographical memory (AM) system andconsiders the process of autobiographical remembering. This model, whichproposes that memories are transitory mental constructions compiled fromdifferent types of autobiographical knowledge, is then applied to the notion ofmemory recovery. The chapter also discusses conscious states associated withthe construction of true and false memories, and it is proposedthat states of conscious awareness characteristic of the two types of memoryconstruction might provide important clues to memory accuracy.
A patient with selective bilateral damage to the amygdala did not acquire conditioned autonomic responses to visual or auditory stimuli but did acquire the declarative facts about which visual or auditory stimuli were paired with the unconditioned stimulus. By contrast, a patient with selective bilateral damage to the hippocampus failed to acquire the facts but did acquire the conditioning. Finally, a patient with bilateral damage to both amygdala and hippocampal formation acquired neither the conditioning nor the facts. These findings demonstrate a double dissociation of conditioning and declarative knowledge relative to the human amygdala and hippocampus.
This chapter explores a question critical to the remainder of the chapters in the title: ‘what are memories?’ It discusses recovered memories and false memories, outlines the recovered memory debate and research into the area, before summarizing the content of the remaining chapters.