Journal of Traumatic Stress, Vol. 11, No.

3, 1998

Predicting Response to Exposure Tkeatment in

PTSD: The Role of Mental Defeat and
Alienation
Anke Ehlers,'~~
David M. Clark', Emma Dunmore,' Lisa Jaycox:
Elizabeth Meadows: and Edna B. Foa2

Not all patients with posttraumatic stress disorder benefit from exposure treat-
ment. The present paper describes two cognitive dimensions that are related
to inferior response to exposure in rape victims. First, individuals whose memo-
ries during reliving of the trauma reflected mental defeat or the absence of
mental planning showed little improvement. Second, inferior outcome was cor-
related with an overall feeling of alienation or permanent change following the
trauma. These results are based on blind ratings of transcripts of exposure treat-
ment sessionsfiom 10 women with good outcome and I0 women with inferior
outcome. Patients in the two groups were matched for initial synptom severiy
and were comparable in many aspects of the assault. Patients who experienced
mental defeat, alienation, or permanent change may require cognitive restnic-
turing in addition to exposure.
KEY WORDS: posttraumatic stress disorder; exposure treatment; predictors of outcome;
mental defeat; alienation.

Many behavioral treatments of posttraumatic stress disorder (PTSD)

involve some form of imaginal reliving of the traumatic event (e.g.,
Boudewyns, Hyer, Woods, Harrison, & McCranie, 1990; Keane, Fairbank,

This manuscript was accepted under the editorial tenure of Bonnie L. Green.
'Department of Psychiatry, University of W o r d , U.K.
'Department of Psychiatry, The Center for the Treatment and Study of Anxiety, Allegheny
University and Hahnernann University, Philadelphia, Pennsylvania 19129.
%o whom correspondence should be addressed at Department of Psychiatry, University of
Oxford, Warneford Hospital, Oxford OX3 7JX U.K. e-mail: ANKE.EHLERS @
PSYCHIATRY.OX.AC.UK.

457

0894-9867/98B7M)-o457S15.w/1 le 1998 Inteinational Society for Traumatic Stress Studies

458 Ehlers, Clark, Dunmore, Jaycox, Meadows, and Foa

Cadell, & Zimering, 1989) and it is widely believed that reliving is an im-
portant component of treatment. The exposure treatment developed by Foa
and colleagues (Foa, Rothbaum, Riggs, & Murdock, 1991) represents the
most systematic application of this therapeutic principle. Controlled trials
have demonstrated that exposure treatment is an effective treatment for
PTSD (Keane et al., 1989; Richards, Lovell, & Marks, 1994) and is more
effective than supportive counselling, and in the long-term, also more ef-
fective than stress inoculation training (Foa et al., 1991).
Although exposure treatment is effective, not all patients benefit from
this approach. For example, Foa et al. (1991) found that of the patients
treated with exposure 45% continued to meet diagnostic criteria for PTSD
and 44% failed to achieve clinically significant improvement at 3-month
follow-up. This raises the question of whether it is possible to identify fac-
tors that predict treatment outcome. The analysis of reasons for inferior
treatment response may also help in developing more effective treatment
procedures.
The present study was designed to explore cognitive differences be-
tween individuals who respond well to exposure and those who show less
marked improvement. Two clinical observations guided the search for pos-
sible predictors. First, we had observed that women who had been raped
differed widely in their reported thought processes during the trauma (see
also Burgess & Holmstrom, 1976). Some reported that they continued to
plan in their own mind how they could influence the situation (even though
these plans were often not successful) and retained a sense of psychoIogica1
autonomy despite the fact that the assailant could physically overpower
them. In contrast, others reported that they felt completely defeated during
the rape and did not feel they were a human being any longer. We expected
that while reliving the former experience should be therapeutic, repeatedly
reliving the latter experience without any cognitive restructuring may not
be.
Second, some patients report post-rape experiences that they perceived
to be negative. These experiences sometimes cause significant distress and
may be traumatic in their own right. We expected poorer outcome for these
patients because exposure treatment does not usually address post-trauma
experiences.
To assess the relationship of thought processes during trauma and post-
trauma experiences with treatment outcome, we obtained narratives of the
traumatic event (rape) from exposure treatment sessions of patients with
good and with inferior outcome. Based on a manual developed for the
purposes of this study, transcripts of these narratives were analyzed by rat-
ers who were blind to treatment outcome. Previous studies have also used
narratives to study variables related to treatment outcome. In contrast to
Mental Defeat and Alienation 459

the study reported here, the focus of those investigations has been on struc-
tural aspects of the narratives such as fragmentation of memories, and not
on the content (see Foa, Molnar, & Cashman, 1995, for results and a re-
view).

Method

Participants

Twenty women who had been sexually assaulted and received exposure
treatment in the context of a study investigating the efficacy of cognitive-
behavioral treatment for PTSD (Foa et al., 1994) comprised the sample
for the present study. Participants with good and inferior outcome were
selected on the basis of their PTSD symptoms at the beginning and end
of treatment, as assessed with the PTSD Symptom Scale (PSS; Foa, Riggs,
Dancu, & Rothbaum, 1993). Groups were matched for initial PSS scores.
Ten participants showed large improvement with treatment (good outcome
group); the other ten showed only modest improvement (inferior outcome
group). Half of the subjects in each group had received prolonged exposure
only (PE, seven sessions), the other half received stress inoculation training
in addition to six sessions of prolonged exposure (PE/SIT). Table 1 shows
participant and assault characteristics for the two groups. The groups did
not differ in age, ethnic background, marital status, education, employment,
pretreatment PSS, depression (Beck Depression Inventory, BDI, Beck,
Ward, Mendelsohn, Mock, & Erbaugh, 1961), anxiety (State-Trait Anxiety
Inventory, STAI-trait and STAI-state, Spielberger, Gorsuch, & Lushene,
1970), months since the assault, percentage of participants who thought
that they would be killed during the assault, percentage of assaults in which
a weapon was used, percentage completed rapes, relationship to assailant,
previous treatment, percentage of participants who reported child sexual
abuse, previous or subsequent rape, or treatment expectancy ratings. The
only significant difference between the groups was that the inferior out-
come group had experienced longer assaults than the good outcome group,
t (10.55) = 3.22, p = .009. As intended by the selection procedure, the
good outcome group showed a much larger reduction in PSS scores with
treatment than the inferior outcome group, Group x Time interaction, F
(1,18) = 11.71, p = .003. There were no effects of treatment condition
(PE vs. PE/SIT) or interactions. On the BDI and STAI, both the good and
inferior outcome groups improved with treatment, time effects, F (1,18) =
23.91, p < .001 for the BDI, F (1,18) = 12.78, p = .002 for STAI-trait, F
(1,18) = 6.88, p = .017 for STN-state.
460 Ehlers, Clark, Dunmore, Jaycox, Meadows, and Foa

Table 1. Participants and Assault Characteristics

Good Inferior
Outcome Outcome
(n = 10) (n = 10)

Age 36.5 (12.2) 32.2 (9.1)

Pre PSS 30.4 (10.5) 32.2 (8.5)
Post PSS 4.0 (3.2) 19.5 (3.7)
Pre BDI 20.3 (12.9) 23.4 (14.4)
Post BDI 6.2 (4.8) 15.3 (10.7)
Pre STAI-trait 51.7 (15.5) 59.2 (10.7)
Post STAI-trait 39.5 (15.0) 49.4 (13.5)
Pre STAI-state 47.9 (15.8) 54.8 (11.5)
Post STAI-state 35.4 (16.4) 46.2 (10.7)
Ethnic group
Caucasian 80% 60%
Black 20% 40%
Marital status
Single 50% 50%
Married 10% 10%
Co-habitating 0% 10%
Divorcedseparated 40% 30%
Employment
Not working 30% 50%
Part time 20% 0%
Full time 50% 50%
Education
High school 10% 50%
College 70% 30%
Graduate school 20% 20%
Months since assault 37.2 (43.6) 107.4 (133.5)
Median: 22.5 Median: 40
Duration of assault (min) 16.9 (11.0) 56.5 (37.3)
% Participants with completed rape 80% 100%
% Participants who thought they would be killed 80% 80%
% Assaults with weapon 50% 60%
Assailant
Stranger 80% 40%
Acquaintance 10% 50%
Relative 0% 10%
Other 10% 0%
% Participants with previous treatent 50% 50%
% Participants reporting child sexual abuse 30% 20%
% Participants with previous rape/attempted rape 0% 30%
% Participants raped again subsequently 10% 10%
Expectancy-treatment logical 7.5 (1.3) 8.4 (1.1)
Expectancy-confident treatment successful for self 7.3 (1.1) 7.2 (1.5)
Expectancy-confident treatment successful for others 6.6 (2.0) 7.5 (1.7)
Mental Defeat and Alienation 461

Materials

Transcripts were obtained for each participant for one early and one
late exposure session. Anxiety ratings and any other possible indicators of
treatment outcome were deleted from the transcripts. The reason for
choosing one early and one late treatment session was that participants
vaned widely in how much information they included in the narratives of
the trauma. The postrape events in particular were sometimes only included
in either the early or the late session. Each session included between two
and three relivings.

Raters

Raters were two Ph.D. level clinical psychologists and two B.A. level
psychologists. The raters were blind with respect to treatment outcome.

Rating Manual

A 10-page manual developed by A.E., D.M.C., and E.B.E described

criteria for the rating scales. Raters were instructed to read the transcripts
once to familiarize themselves with the stoq. They then read the transcript
again highlighting any material relevant to the concept of mental planning
vs. mental defeat, and overall feeling of alienation/permanent change. The
evidence was then summarized on a separate sheet of paper. After review-
ing the evidence from the transcripts of the two treatment sessions, raters
gave the overall ratings.
The bipolar concept “mental planning vs. mental defeat” refers to the
victims’ thoughts during the assault. Mental planning (coping) is defined as
thinking about or planning in one’s mind about what she might be able to
do to minimize physical or psychological harm, to make the experience
more tolerable, or to influence the response of the assailant (irrespective
of whether the plan is successful or not). Evidence for mental planning is
taken from planning thoughts and statements that a particular action was
taken intentionally. Mental defeat refers to the victim’s perception that she
gave up in her own mind and was completely defeated. Evidence for mental
defeat is taken from direct statements that the victim gave up, her descrip-
tion of feeling completely at the will of the assailant, her description that
she lost the sense of being a person with her own will, statements that she
did not feel like a human being any longer, her wisWacceptance that she
was dying, or a sexual response during the rape which she perceived as
defeat. Mental defeat is rated conservatively and goes beyond helplessness
462 Ehlers, Clark, Dunmore, Jaycox, Meadows, and Foa

and hopelessness. Even if a victim cannot do anything to prevent the rape

she may retain a sense of her own will and of being psychologically inde-
pendent of the assailant’s will.
Evidence for mental planning and mental defeat is summarized in an
overall rating scale with “0” meaning strong evidence for mental planning
and no evidence for mental defeat, “1” meaning some evidence for mental
planning and no evidence for mental defeat, “2” meaning neither evidence
for mental planning nor mental defeat, and “3” clear evidence for mental
defeat, irrespective of whether or not mental planning was present at other
times. Pairwise interrater-reliabilities (Pearson correlations, n = 20 ratings)
between the four raters ranged from .73 to -90 for the Mental PlanningDe-
feat scale (mean: 33).
The concept of “overall feeling of alienation/permanent change” refers
to negative post-rape experiences. Patients who score high on this scale
report that they could not return to their normal lives after the rape. They
felt alienated from themselves, their world, and their future. Evidence for
an overall feeling of alienation is taken from the victim’s description of her
interaction with others after the assault and her feelings about these inter-
actions. The types of experiences that are likely to give the victim an overall
feeling of alienation are, e.g., being blamed, being mistreated, or not being
believed. These statements are only taken as evidence if it is clear that
victims experienced them as negative and did not discount the person as
not relevant, not competent, etc. Further examples for experiences relevant
to the concept of alienation are the victim’s report that she was unable to
relate to other people after the assault or that other people did not meet
her needs. Statements or descriptions which indicate that the victim felt
supported or safe with other people after the rape are rated as evidence
against an overall feeling of alienation. Higher ratings on overall feeling
of alienation reflect the impression that the victim lost the sense of being
supported by her social network and of herself as a member of the network.
The rating of an overall feeling of alienation is not identical with the DSM-
criterion and PSS-item “feeling detached and estranged from others” be-
cause it requires the victim to feel alienated in general, not just in
relationship to some people. Furthermore, the Overall Alienationperma-
nent Change rating was concerned with immediate post-trauma experiences
whereas the PSS assesses current symptoms. The correlation between the
Overall Alienationpermanent Change scale and the PSS detachment item
was nonsignificant (rho = -.17). Evidence for permanent change (damage)
is taken from statements indicating that victims see themselves or their lives
as having been changed forever in a negative way, e.g., statements convey-
ing a sense that life is over, a sense that they have been damaged forever,
a sense that they died inside, a sense that their body has been permanently
Mental Defeat and Alienation 463

damaged or spoilt, or a sense that relationships with others will never be

the same again. Statements that indicate that the victim saw the assault as
an isolated bad experience are rated as evidence against permanent change.
Evidence on negative post-assault experiences is summarized in a rat-
ing scale ranging from “0” meaning no evidence to “4” meaning very strong
evidence for an overall feeling of alienatiodpermanent change (with “1”
meaning weak evidence, “2” moderate evidence and “3” strong evidence).
Pairwise interrater-reliabilities between the four raters for the Overall Al-
ienatiooermanent Change scale ranged from .46 to .SO (mean: .66, n =
19 ratings). Reasons for discrepancies were twofold: First, raters identified
the same experiences as relevant but gave them a different weight. This
rating was especially difficult because sometimes there was only one reliving
phase in which post-rape experiences were included. Second, raters some-
times missed evidence discounting a person who had acted in a negative
way towards the victim as not relevant (e.g., “they are idiots,” “they were
not doing their job properly”). In these cases they overrated alienation.

Results

Relationship of Treatment Outcome to Mental PlanninglDefeat and Overall

AlienationlPermanent Change

Table 2 shows the mean ratings of the four raters for the good and
inferior outcome groups. Participants with inferior outcome received higher
scores on the Mental Planning/Defeat scale than participants with good
outcome, t (11.59) = 4.06, p = -002. The inferior outcome group showed
very little evidence of mental planning and approximately half experienced
mental defeat. None of the participants with good outcome were rated as
having experienced mental defeat and all of them reported evidence of
mental planning. There was a significant negative correlation between the
Mental Planning/Defeat scale and percent improvement in the PSS during
treatment, r(18) = -.66,p = .002.

Table 2. Mean Ratings of Four Raters Blind to Treatment Outcome

Good Inferior
Outcome Outcome
Mental planning vs. defeat (0-3) 0.45 (0.41) 1.90 (1.06)
Overall feeling of alienatiodpermanent change (0-4) 1.65 (1.16) 2.65 (1.12)
464 Ehlers, Clark, Dunmore, Jaycox, Meadows, and Foa

For the Overall Alienatioflermanent Change scale, a trend for a sig-

nificant group difference was observed, t (17) = 1.91, p = .07. The scale
correlated negatively with the percent improvement in PTSD symptoms,
417) = -.48,p = .04.
The inferior and good outcome groups differed in duration of assault,
which in turn correlated with Mental PlanningDefeat, r = 59. p =.006.
This raised the possibility that assault duration might account for the re-
lationship between Mental Planninmefeat and outcome. Analysis of co-
variance (ANCOVA) excluded this possibility as the inferior outcome group
continued to score higher on Mental PlanninglDefeat when assault duration
was covaried out, F(1,17) = 6.31, p = .02. Relationship to the rapist
(stranger vs. known) also correlated with Mental PlanninglDefeat, r = .65,
p =.002, such that those raped by strangers showed more mental planning
and less mental defeat. Another ANCOVA tested whether this variable
could explain the relationship between Mental PlanningDefeat and out-
come. Once again the group difference in Mental PlanninglDefeat re-
mained significant when this potentially confounding variable was covaried
out, F (1,17) = 10.12, p = .005. ANCOVAs were not performed on the
Overall Alienatioflermanent Change data as this variable was not corre-
lated with assault duration or relationship to the rapist.

Relationship of Other ‘Variables to Treatment Outcome

Table 3 shows the correlations between percent improvement in PSS

score and each of the patient/assault characteristics in Table 1. As expected
from the group analysis, assault duration was significantly negatively cor-
related with outcome. There were also trends for attempted vs. completed
rape and the PSS-item “detachmendestrangement from others” to correlate
with outcome. Neither of the latter variables correlated with Mental Plan-
ning/Defeat (rho = .18 and .15) or Overall Alienatioflermanent Change
(rho = .ll and -.17).

Discussion

The results obtained with blind ratings of patients’ recollection of the

assault confirmed our clinical impression that patients’ thoughts during
rape as well as their perception of post-rape experiences are related to
their response to exposure treatment. Inferior outcome was related to pres-
ence of mental defeat or absence of mental planning. Patients who did not
plan or who gave up thinking about how they could minimize harm during
Mental Defeat and Alienation 465

Table 3. Correlations between Patient and Assault Characteristics and

Treatment Outcome@
~~~ ~

Percentage
change in PSS

Age .13
Months since assault (rho) -.I6
Duration of assault (r) -.49*
Ethnic group (Black vs. Caucasian, rho) .11
Marital status (singleldivorced vs. marriedco-habitating, rho) -.30
Employment (rho) .06
Years of education ( r ) .I5
Depression (BDI pre, r ) .06
Detachment from others (PSS item 9, rho) -.42"
Attempted vs. completed rape (rho) -.40**
Expectation to be killed (no vs. yes, rho) -.09
Presence of weapon during assault (rho) .06
Assailant stranger vs. known (rho) -.35
Previous treatment (rho) -.02
Child sexual abuse (no vs. yes, rho) .27
Previous rape/attempted rape (no vs. yes, rho) -.35
Subsequent rape (rho) -.26
Eupectancy-treatment logical ( r ) -.32
Expectancy-confidence self ( r ) .I6
Expectancy-confidence others ( r ) -.14
'y: Pearson correlation; rho: Spearman correlation; ' p < .05; **p < .lo.

the rape and especially those who felt completely defeated and lost the
sense of being a person with their own will responded less well to repeated
imaginal reliving of the traumatic event.
One may argue that giving up planning or experiencing mental defeat
during rape may reflect differences in other comorbid symptoms such as
depression, objective characteristics of the assault, or be due to previous
negative experience such as child sexual abuse. However, a large number
of possibly relevant patient and assault characteristics were assessed, and
the only significant difference between the groups was in assault duration.
Although the experience of mental defeat was correlated with duration of
the assault and relationship to assailant, its relationship with treatment out-
come could not fully be explained by these variables. These results make
it highly unlikely that the cognitive differences between the good and in-
ferior outcome groups were simply epiphenomena of objective differences
in trauma history. The analysis of correlation patterns with percent im-
provement in posttraumatic symptoms underscores this conclusion.
When interpreting the low correlations between assault and previous
trauma characteristics with PSS change scores, one has to bear in mind
that all the patients in the sample had experienced a very severe trauma:
466 Ehlers, Clark, Dunmore, Jaycox, Meadows, and Foa

the vast majority of the participants thought they would be killed, and
nearly all experienced completed rape. The impact of pre-rape experiences
and assault characteristics would probably be larger in samples with a wider
range of trauma severity. In addition, although many of the correlations
were nonsignificant, some were in the expected direction of a relationship
between overall severity of trauma history (e.g., previous and subsequent
rape) and outcome. It is possible that participants who had been raped
repeatedly and those who experienced longer assaults may have benefited
from a larger number of exposure sessions. However, the pattern of results
may also point to the need to distinguish between factors predicting the
initial occurrence of PTSD symptoms and those predicting chronicity of
symptoms or poor response to treatment (Ehlers & Steil, 1995; Rothbaum,
Foa, Riggs, Murdock, & Walsh, 1992).
There are several ways in which mental defeat/absence of mental plan-
ning may undermine the efficacy of exposure therapy. The experience of
mental defeat may impede recovery by preventing the patient from seeing
that the trauma was a single, time-limited, past event which does not nec-
essarily have global implications for oneself or one’s future. Reliving the
event in exposure therapy may be less effective and possibly retraumatizing
because the patient is repeatedly reminded of hidher inability to cope. This
may confirm some negative views the victim holds about herself (e.g., “I
am worthless,” “I deserved it,” “I am to blame for what happened,” “I am
disgusting,” “I am a bad person”) and about her ability to cope with future
adverse events or the symptoms triggered by the assault (e.g., “I am a lousy
coper,” “It will happen again and I will not be able to prevent it,” “Bad
things will always happen to me”). Reliving may thus fail to disconfirm the
belief that the traumatic event has serious implications for the future, and
maintain rather than change the patient’s sense of ongoing threat. In line
with our results, Pitman et al. (1991) observed that six veterans with PTSD
who experienced complications such as depression or relapse of alcoholism
during exposure treatment were characterized by negative appraisal of their
actions during the combat events they were exposed to. Repeated reliving
of the combat events seemed to confirm rather than change these negative
appraisals.
This raises a general issue about the use of exposure therapy in PTSD.
According to Foa and Kozak (1986) and Foa, Steketee, and Rothbaum
(1989), exposure therapy leads to the incorporation of new corrective in-
formation into the fear network. Foa and Jaycox (in press) have outlined
several ways in which this might occur in patients with chronic PTSD. Re-
peated reliving of the trauma lowers anxiety by habituation. Reliving of the
trauma in a safe environment incorporates safety information into the trau-
matic memory. Facing the trauma helps patients to discover they will not
Mental Defeat and Alienation 467

go mad or lose control and to distinguish the memory from present reality.
Finally, traumatic memories tend to be disorganized and repeated reliving
helps patients to construct a coherent narrative which assists in putting the
experience behind them. Exposure therapy was initially developed for the
treatment of phobias and obsessive-compulsive disorder. It may apply best
to those symptoms and cognitions connected with PTSD that most closely
resemble those in these disorders, namely those related to excessive per-
ception of danger. For example, a woman who was raped may think after-
wards “All men are dangerous.” Exposure may help change this cognition
in a way similar to changing a phobic’s belief “All dogs are dangerous.”
However, exposure may be less effective at changing cognitions related to
negative self-evaluation such as those targeted in this study. These cogni-
tions may need to be addressed directly with cognitive therapy procedures.
It remains to be investigated whether individuals who experienced mental
defeat during rape will respond better to a combination of exposure and
cognitive restructuring, to cognitive restructuring alone, or to a larger num-
ber of exposure treatment sessions than used in the present study. We
would hypothesize that the combination of these methods will be most ef-
fective because reliving the event facilitates the assessment and modifica-
tion of the idiosyncratic meaning of the trauma, and cognitive restructuring
may be most effective in changing cognitions related to negative self-evalu-
ation.
One may argue that the relationship between mental planning and
treatment outcome reflects the role of perceived uncontrollability which is
a crucial variable in determining the aversiveness of a negative event (e.g.,
Foa et al., 1989; Foa, Zinbarg, & Rothbaum, 1992; Mineka, 1985). Mental
planning can be understood as attempts to exert control over the situation
(see also Burgess & Holmstrom, 1976), even if the control was minimal or
symbolic, such as a patient not wanting to show the rapist that she was
crying. Lack of mental planning may therefore reflect the patients’ percep-
tion that the situation is totally uncontrollable. However, mental defeat may
go beyond uncontrollability. In the present study, there was no difference
between the groups in their ability to prevent the rape and 80% were con-
vinced they would be killed during the rape. All patients described feeling
helpless, hopeless, and humiliated. The difference seemed to be that those
with good outcome continued to retain a sense of autonomy. The impor-
tance of retaining a sense of personal autonomy may explain Resnick, Kil-
patrick, Dansky, Saunders, and Best’s (1993) finding that completely
uncontrollable events like natural disasters are less likely to lead to PTSD
than trauma inflicted by other people such as rape, physical assault, or
torture.
468 Ehlers, Clark, Dunmore, Jaycox, Meadows, and Foa

It is interesting to note that DSM-IV (American Psychiatric Associa-

tion [MA],1994) emphasizes the role of perceived threat to physical in-
tegrity in the development of PTSD. The two groups were equated on this
variable. The role of mental defeat in predicting treatment response in the
present study as well as persistence of PTSD in a study comparing patients
with persistent vs. recovered PTSD (Dunmore, Clark, & Ehlers, 1997) sug-
gest that threat to psychological integrity may be of additional importance.
Results for Overall Alienatioflermanent Change were less clear-cut
but nevertheless pointed in the expected direction. There was a trend for
a difference between the good and inferior outcome groups and the scale
correlated significantly with degree of change in posttraumatic symptoms.
Patients whose recollected postrape experiences indicated an overall feeling
of alienation from self or others, or permanent negative change, benefited
less from exposure to their rape memories than patients without these char-
acteristics. These results were not a reflection of an overlap between Over-
all Alienatioflennanent Change and the PTSD symptom “detachment
and estrangement from others” because there was no correlation between
these measures. These aspects of alienation were independently related to
treatment outcome. The present study may underestimate the true effects
of Overall Alienation/F’ermanent Change because little relevant material
was available for each patient and some patients may not have volunteered
relevant information because they were not asked to talk about postrape
experiences during reliving.
Several mechanisms may underlie the relationship between Overall Al-
ienatioflermanent Change and inferior treatment response. First, the re-
living used in this study mainly focused on the traumatic event and paid
little attention to post-rape experiences. Therapists often interrupted pa-
tients when their reliving of the rape was finished and asked them to start
again. As a consequence, treatment did not directly address patient per-
ceptions of Overall Alienatioeermanent Change. Second, some distress-
ing memories of the rape which occur during exposure may be interpreted
by some patients as further evidence that they are permanently changed.
Third, alienation from others may be linked to the patients’ preoccupation
with feelings such as guilt, anger, or shame. Prospective studies of post-
traumatic symptoms after trauma have shown that preoccupations with
these feelings are related to inferior long-term outcome (Amir, Foa, &
Cashman, 1996; Riggs, Dancu, Gershuny, Greenberg, & Foa, 1992) and
there is other evidence that they are linked to poor treatment response
(Foa, Riggs, Massie, & Yarcower, 1995; Jaycox & Foa, 1996; Vaughan &
Tarrier, 1992). There are several possible explanations for the role of pre-
occupations with guilt and anger. Ehlers and Steil (1995) have argued that
these preoccupations are functionally similar to avoidance in that they pre-
Mental Defeat and Alienation 469

vent exposure to the emotions experienced during the trauma. Riggs et al.
(1992) have proposed that the activation of anger networks inhibits fear
responses which would be necessary to achieve changes in the fear network
representing the trauma.
In line with our finding of a relationship between alienation and treat-
ment outcome, several recent studies have obtained results suggesting post-
trauma social interactions that are perceived as negative hinder post-trauma
recovery in untreated victims of trauma. Davis, Brickman, and Baker (1991)
and Ullman (1996) found that reports of negative social interaction were
associated with poor adjustment. Dunmore et al. (1997) found that negative
appraisals of other people’s responses after assault distinguished between
individuals with persistent versus recovered PTSD.
The material used for the current study does not allow one to decide
whether patients who experienced alienation actually received more nega-
tive responses or less support from other people than patients who did not
feel alienated. However, the coding instructions for alienation emphasized
patients’ interpretations of others’ responses. Some patients seemed to be
unaffected by negative responses because they discounted them as not rele-
vant whereas others were very upset about similar instances. Furthermore,
some patients felt alienated because of sympathetic responses which they
interpreted as other people looking down on them. It is therefore likely
that the patients’ interpretation of others’ responses plays a crucial role. It
is also likely that beliefs patients hold about what response they can expect
from others will actually have an impact on their social interactions in that
it will influence their behavior and thus the response they will receive from
other people. When interpreting the results of the present study, one has
to bear in mind that the inferior outcome group did not represent treat-
ment failures because they also showed improvement with exposure. Com-
pared to the good outcome group, however, their progress was much less
impressive. Thus, the predictor variables should be interpreted as impeding
rather than preventing therapeutic change during exposure.

Acknowledgements

The research described in this paper was funded by the Wellcome Trust
and a grant from the National Institute of Mental Health (MH42178). We
would like to thank Chris Molnar and Jane Stafford for their help with
data collection, Karen Bogert for helping produce the transcripts, and
Sarah Durbin and Fiona Lobban for rating the transcripts.
470 Ehlers, Clark, Dunmore, Jaycox, Meadows, and Foa

References

American Psychiatric Association (1994). Diagnostic and statistical manual of mental disorders
(4th ed.). Whington, D.C.: Author.
Amir, N., Foa, E. B., & Cashman, L. (1996). A predictive model of posttraumatic stress
disorder. Paper in preparation.
Beck, A. T, Ward, C. H., Mendelsohn, M.. Mock, J., & Erbaugh, J. (1961). An inventory for
measuring depression. Archives of General PsychiaQ, 4, 561-571.
Boudewyns, I? A. Hyer, L., Woods,M. G., Harrison, W. R., & McCranie, E. (1990). PTSD
among Vietnam veterans: An early look at treatment outcome using direct therapeutic
exposure. Joumal of Traumatic Stress, 16, 29-33.
Burgess, A. W.,& Holmstrom, L. L. (1976). Coping behavior of the rape victim. American
Journal of Psychiatg 133, 413-418.
Davis, R. C., Brickman, E. R., & Baker, T (1991). Effects of supportive and unsupportive
responses of others to rape victims: Effects of current victim adjustment. American Journal
of Community Psychology, 19, 443-451.
Dunmore, E., Clark, D. M., & Ehlers, A. (1997). Cognitive factors in persistent vs. recovered
post-traumatic stress disorder after physical or sexual assault: A pilot study. Behavioural
and Cognitive Psychotherapy, 25, 147-159.
Ehlers, A., & Steil, R. (1995). Maintenance of intrusive memories in posttraumatic stress
disorder: A cognitive approach. Behavioural and Cognitive Psychotherapy, 23, 217-249.
Foa, E. B., Freund, B. E, Dancu, C. V, Franklin, M. E., Perry, K J., Riggs, D. S., & Molnar,
C. (1994,November). EfJcancy of shod-tern behavioral treatment of PTSD in seruaf and
non-sexual victims. Paper presented at the Annual Meeting of the Association for the
Advancement of Behavior Therapy, San Diego, California.
Foa, E. B., & Jaycox, L H. (in press). Cognitive-behavioral treatment of post-traumatic stress
disorder. In D. Spiegel (Ed.), The practice of psychotherapy. Washington, D C American
Psychiatric Press.
Foa, E. B., & Kozak, M. J. (1986). Emotional processing of fear: Exposure to corrective
information. Psychological Bulletin, 99, 20-35.
Foa, E.B., Molnar, C., & Cashman, L. (1995). Change in narratives during exposure therapy
for posttraumatic stress disorder. Journal of Faumatic Stress, 8, 675-690.
Foa, E. B., Riggs, D. S., Dancu, C. V, & Rothbaum, B. 0. (1993). Reliability and validity of
a brief instrument for assessing post-traumatic stress disorder. Journal of Traumatic Stress.
6, 459-473.
Foa, E. B., Riggs, D. S., Massie, E. D., & Yarczower, M. (1995). The impact of fear activation
and anger on the efficacy of exposure treatment for PTSD. Behavior Therapy, 26,487-500.
Foa, E. B., Rothbaum, B. O., Riggs, D. S., & Murdock, T B. (1991). Treatment of
posttraumatic stress disorder in rape victims: A comparison between cognitive-behavioral
procedures and counseling. Journal of Consulting and Clinical Psychology, 59, 715-723.
Foa, E. B., Steketee, G.. & Rothbaum, B. 0. (1989). Behavioralkognitive conceptEa1izations
of post-traumatic stress disorder. Behavior Therapy, 20, 155-176.
Foa, E. B., Zinbarg, R.. & Rothbaum, B. 0. (1992). Uncontrollability and unpredictability in
post-traumatic stress disorder: An animal model. Psychological Bulletin. 112. 218-238.
Jaycox, L. H.,& Foa, E. B. (1996). Obstacles in implementing exposure therapy for PTSD:
Case discussions and practical solutions. Clinical Psychology and Psychotherapy, 3, 176-184.
Keane, T M., Eairbank, J. A., Caddell, J. M., & Zimering, R. 'I: (1989). Implosive (flooding)
therapy reduces symptoms of PTSD in Vietnam combat veterans. Behavior i%erapy, 20,
245-260.
Mineka, S. (1985). Animal models of anxiety-based disorders. In A. H. Tuma & J. D. Maser
(Eds.), Anriey and the am'dy disorders (pp. 199-244). Hillsdale, NJ: Erlbaum.
Pennebaker, J. W. (1989). Confession, inhibition, and disease. Advances in Experimental Social
PSychOlOgY, 22, 211-244.
Mental Defeat and Alienation 471

Pitman, R. K., Altman, B., Greenwals, E., Longpre, R. E., Macklin, M. L., Poire, R. E., &
Steketee, G. S. (1991). Psychiatric complications during flooding therapy for posttraumatic
stress disorder. Journal of Clinical Psychiairy, 52, 17-20.
Spielberger, C. D.. Gorsuch, R. L,& Lushene. R. E. (1970). Stare-Trait Anxieiy Inventory
(Self-evaluationQuestionnaire). Palo Alto, C A Consulting Psychologists Press.
Resnick, H. S., Kilpatrick, D. G., Dansky, B. S., Saunders, B. E., & Best, C. L. (1993).
Prevalence of civilian trauma and posttraumatic stress disorder in a representative
national sample of women. Journal of Consuliing and CIinical Psychology, 61, 984-991.
Richards. D. A., Lovell, K, & Marks, I. M. (1994). Posttraumatic stress disorder: Evaluation
of a behavioral program. Journal of Traumatic Stress, 7, 669-680.
Riggs, D. S., Dancu, C. V, Gershuny, B. S., Greenberg, D., & Foa, E. B. (1992). Anger and
post-traumatic stress disorder in female crime victims. Journal of Eaumatic Stress, 5,
613-625.
Rothbaum, B. 0.. Foa, E. B., Riggs. D. S., Murdock, T,& Walsh, W. (1992). A prospective
examination of posttraumatic stress disorder in rape victims. J O U ~ ofQ ~Traumatic Siress,
5, 455-475.
Ullman, S. E. (1996). Correlates and consequences of adult sexual disclosure. Journal of
Interpersonal Kolence, I I , 554-571.
Vaughan, K., & Tarrier, N. (1992). The use of image habituation training with posttraumatic
stress disorder. Briiish Journal of Psychiatry, 161, 658-664.