Prolonged Exposure versus Eye Movement Desensitization and Reprocessing (EMDR) for PTSD rape victims

Article (PDF Available)inJournal of Traumatic Stress 18(6):607-16 · December 2005with3,856 Reads
DOI: 10.1002/jts.20069 · Source: PubMed
This controlled study evaluated the relative efficacy of Prolonged Exposure (PE) and Eye Movement Desensitization and Reprocessing (EMDR) compared to a no-treatment wait-list control (WAIT) in the treatment of PTSD in adult female rape victims (n = 74). Improvement in PTSD as assessed by blind independent assessors, depression, dissociation, and state anxiety was significantly greater in both the PE and EMDR group than the WAIT group (n = 20 completers per group). PE and EMDR did not differ significantly for change from baseline to either posttreatment or 6-month follow-up measurement for any quantitative scale.


Journal of Traumatic Stress, Vol. 18, No. 6, December 2005, pp. 607–616 (
Prolonged Exposure Versus Eye Movement Desensitization
and Reprocessing (EMDR) for PTSD Rape Victims
Barbara Olasov Rothbaum,
Millie C. Astin,
This controlled s tudy evaluated the relative efficacy of Prolonged Exposure (PE) and Eye Movement
Desensitization and Reprocessing (EMDR) compared to a no-treatment waitlist control (WAIT) in
the treatment of PTSD in adult female rape victims (n = 74). Improvement in PTSD as assessed by
blind independent assessors, depression, dissociation, and state anxiety was significantly greater in
both the PE and EMDR group than the WAIT group (n = 20 completers per group). PE and EMDR
did not differ significantly for change from baseline to either posttreatment or 6-month follow-up
measurement for any quantitative scale.
Posttraumatic stress disorder (PTSD) is a major
health problem in the United States, in that as many as
one half of rape victims may suffer from chronic PTSD
(Rothbaum, Foa, Riggs, Murdock, & Walsh, 1992). As
nearly 12 million American women have a lifetime history
of PTSD because of rape (Resnick, Kilpatrick, Dansky,
Saunders, & Best, 1993), rape victims may constitute the
largest number of PTSD sufferers in the United States;
therefore, it is imperative to study effective and efficient
treatments for rape victims with this disorder.
In the treatment guidelines for PTSD psychothera-
pies, prepared under the auspices of the International So-
ciety for Traumatic Stress Studies (ISTSS), the strongest
evidence was found for the use of cognitive behavioral
techniques (CBT) to treat PTSD (Foa, Friedman, & Keane,
2000). Of the CBT techniques studied, prolonged imag-
inal exposure (PE) had the most evidence for its effi-
cacy (Rothbaum, Meadows, Resick, & Foy, 2000). Based
on seven controlled studies with large effect sizes, Eye
Movement Desensitization and Reprocessing (EMDR)
Department of Psychiatry and Behavioral Sciences, Emory University
School of Medicine, Atlanta, Georgia.
Behavioral Research Consulting, Inc., Stone Mountain, Georgia.
To whom correspondence should be addressed at Department of Psy-
chiatry and Behavioral Sciences, The Emory Clinic, 1365 Clifton Road,
Atlanta, Georgia 30322; e-mail:
has been deemed to be an efficacious treatment for PTSD
(Chemtob, Tolin, van der Kolk, & Pitman, 2000).
Several published studies have compared PE to
EMDR (Devilly & Spence, 1999; Ironson, Freund,
Strauss, & Williams, 2002; Lee, Gavriel, Drummond,
Richards, & Greenwald, 2002; Rogers, Silver, Goss,
Obenchain, Willis, & Whitney, 1999; Simon, 2000; Taylor
et al., 2003) with mixed results. Most have been method-
ologically weak with small sample sizes ranging from 1
to 12 participants per treatment modality (see Table 1).
Indeed, of the seven gold standards for PTSD treatment
outcome research identified by Foa and Meadows (1997),
only one published study (Taylor et al., 2003) to date met
all seven standards. Taylor et al. (2003) found that PE
was superior to both EMDR and relaxation at posttest
and follow-up, with no differences between EMDR and
relaxation outcomes. One other study (Lee et al., 2002)
met five of the seven standards. Comparing EMDR to
combined Stress Inoculation Training and Prolonged Ex-
posure (SITPE), Lee et al. (2002) found that EMDR did
not differ from SITPE, except on measures of PTSD in-
trusion; however, at follow-up, EMDR led to greater gains
on all measures.
It has been suggested that PTSD occurs due to the
inability to adequately process the trauma (Foa, Steketee,
& Rothbaum, 1989). Foa and Kozak (1986) suggested that
2005 International Society for Traumatic Stress Studies Published online in Wiley InterScience ( DOI: 10.1002/jts.20069
608 Rothbaum, Astin, and Marsteller
Table 1. EMDR Studies Compared on “Gold Standards”
Devilly & Spence, Ironson et al., Lee et al., Rogers et al., Simon, Taylor et al.,
Study 1999 2002 2002 1999 2000 2003
EMDR vs.
EMDR vs.
Exposure vs.
Sample size 23 22 24 12 2 45
Standard 1: clearly
defined symptoms
Yes Yes Yes Yes Yes Yes
Standard 2: reliable
and valid measures
Yes Yes (self-report
measures only)
Yes Yes Yes Yes
Standard 3: use of
No (self-report
Inconsistent Yes No Yes
Standard 4: trained
Not specified No (self-report
Yes Not specified Not specified Yes
Standard 5:
replicable, specific
treatment programs
Yes Yes Yes Yes (but different
therapists for
each treatment)
Not specified Yes
Standard 6: treatment
Yes (but results
not reported)
Yes (but not
Yes (but results
not reported)
No No (therapist
same for both
treatments and
not trained in
Standard 7: unbiased
assignment to
due to PE
Yes Yes Not specified Yes
Note. Used stratified method—first 10 to TTP and second 10 to EMDR, and then random, TTP = Trauma Treatment Protocol (includes prolonged
exposure); SITPE = Stress Inoculation Training Combined With Prolonged Exposure; EMDR = Eye Movement Desensitization and Reprocessing.
two conditions are required for the reduction of fear. First,
the fear memory must be activated. Second, new informa-
tion must be provided that changes the fear structure. Foa
and Kozak proposed that any successful therapy involves
correcting the pathological elements of the fear structure,
and that this corrective process is the essence of emotional
processing. The theoretical foundation behind EMDR is
less well developed at this point. It is likely that EMDR is
functioning as another form of exposure by confronting
the patient with the traumatic memory, thus aiding emo-
tional processing of the traumatic memory. The specific
role of the eye movements is unclear, although they may
serve as a distraction by allowing the patient to remain
with the upsetting image and reactions (Barrowcliff,
Gray, MacCulloch, Freeman, & MacCulloch, 2003;
Devilly, 2001; Lohr, Tolin, & Lilienfeld, 1998; Rogers
& Silver, 2002).
This study aimed to evaluate the relative efficacy of
EMDR versus standard PE compared to a no-treatment
waitlist control (WAIT) in the treatment of PTSD in adult
female rape victims. The specific hypotheses tested in-
clude that (a) both EMDR and PE would be more effective
in reducing PTSD symptoms than WAIT, (b) EMDR and
PE would be equally effective at reducing PTSD symp-
toms at posttreatment, and (c) EMDR participants would
improve more quickly than PE participants.
Eligibility Criteria
This study was limited to female victims of a rape at
least 3 months prior to study entry to allow for the natural
decline in PTSD symptoms (Kilpatrick & Calhoun, 1988;
Rothbaum & Foa, 1989). To obtain a diverse sample of
rape victims with chronic PTSD, no maximum time since
the index rape was imposed. The PTSD assessment was
conducted for the participant’s subjectively most trau-
matic rape event.
Inclusion criteria were that the index event must have
been a rape in adulthood (i.e., age 12 or older) or a sin-
gle incident of rape in childhood (ages 0–11) by either
a family or a nonfamily member. Participants were not
excluded if they had other traumas, including childhood
sexual abuse. Rape was defined as any form of unwanted
genital penetration including vaginal, anal, oral, and dig-
ital penetration. Fondling or touching of genitals through
clothes were not included. Three participants who experi-
enced life-threatening events that included attempted rape
without actual penetration as defined earlier were allowed
to participate. If on psychotropic medication, participants
were required to be s table on the medication at the same
dosage for 30 days prior to study entry and to agree not
PE Versus EMDR for PTSD 609
to change medication or dosage for the duration of the
Exclusion criteria included (a) a history of
schizophrenia or other psychoses, (b) current suicidal risk
or practiced self-mutilation, (c) illiterate and thus unable
to complete self-reports, (d) current alcohol or drug depen-
dence as determined by the Structured Clinical Interview
for the Diagnostic and Statistical Manual of Mental Dis-
ease (DSM-IV; American Psychiatric Association, 1994)
(SCID; First, Gibbons, Spitzer, & Williams, 1996), (e)
blind or had a history of serious eye disease (e.g., detached
retina) that would cause risk with rapid eye movement, (f)
use of cocaine in any form within 60 days of treatment
administration, (g) or in an ongoing threatening situation
(e.g., domestic violence).
Phone screening was conducted by the study coordi-
nator to determine appropriateness of t his study for each
potential participant and to inform her of study proce-
dures. If appropriate, an initial evaluation was scheduled.
At this evaluation, participants were evaluated by an In-
dependent Assessor (IA) with respect to inclusion and ex-
clusion criteria outilined previously, and the IA explained
the procedures of the study to them in detail. Both written
and verbal informed consent were gathered at the pre-
treatment assessment conducted by the IA. If the partici-
pant met criteria and gave consent, she was then random-
ized and scheduled accordingly: Seventy-four participants
were randomly assigned to one of two active treatments
(EMDR or PE) or a waitlist control group (WAIT). The
informed consent form and this study were approved by
the Institutional Review Board.
The Clinician-Administered PTSD Scale (CAPS;
Blake et al., 1995; Blake et al., 1990; Weathers et al.,
1992) is a clinician-administered structured interview that
assesses the 17 DSM-IV PTSD symptoms for both fre-
quency and severity and totaled for the three major symp-
tom clusters (intrusion, avoidance, and hyperarousal). The
CAPS Current and Lifetime Version (1-month symptom-
duration criteria) was used for pretreatment and follow-
up assessments whereas the CAPS One Week Version
(1-week symptom-duration criteria) was used for the
postwaitlist and posttreatment assessments. For the cur-
rent study, interrater reliability for the CAPS was 93.8%
(κ = .79).
The Assault Information Interview (AII; Rothbaum
et al., 1992) is a structured interview that gathers in-
formation during the first treatment session on relevant
aspects of the assault such as acts committed, number
of assailants, weapons used, and so on. Similarly, the
Treatment, Legal, and Drug Update Interview (UPDATE;
Rothbaum, 1997), used at posttreatment and follow-up
assessments, is a structured interview that gathers infor-
mation on other treatment received, legal developments
of the rape case, and alcohol and drug use occurring after
the onset of the study.
The Stressful Life Events Screening Questionnaire
(SLESQ: Goodman, Corcoran, Turner, Yuan, & Green,
1998) is an interview developed to ascertain a compre-
hensive trauma history. Developed recently, the SLESQ
shows good test-retest reliability, adequate convergent va-
lidity with a lengthier clinical interview of prior trauma
history, and good discrimination between Criterion A and
non-Criterion A events.
The SCID Non-Patient Version (First et al., 1996) is
a structured diagnostic interview based on criteria from
the DSM-IV and is used extensively in research and clini-
cal settings. Particular modules administered included the
Mood Disorders, Substance Use Disorders, and Anxiety
Disorders modules to screen for other DSM-IV highly co-
morbid axis I disorders. For the current study, interrater
reliability for the SCID was 96.8% (κ = .83).
Self-Report Measures
The PTSD Symptom Scale-Self-Report (PSS-SR:
Foa, Riggs, Dancu, & Rothbaum, 1993; Rothbaum,
Dancu, Riggs, & Foa, 1990) is a 17-item self-report mea-
sure that corresponds to the 17 DSM-IV symptoms of
PTSD. The presence and severity of PTSD items are
ratedona0(not present)to3(very much) s cale, with
subscores available for reexperiencing, avoidance, and
The Impact of Event Scale-Revised (IES-R: Weiss
& Marmar, 1997) is based on the original IES (Horowitz,
Wilner, & Alvarez, 1979) that was comprised of 15 items
measuring intrusive and avoidance/numbing symptoma-
tology. The IES-R is a 22-item scale that includes 7 new
items to assess the third PTSD symptom cluster, hyper-
arousal. The frequency of each item is rated from 0 (not
at all)to4(extremely).
The Beck Depression Inventory (BDI: Beck, Ward,
Mendelsohn, Mock, & Erbaugh, 1961; Beck, Steer, &
Garbin, 1988) is a 21-item self-report questionnaire
widely used in research on depression to evaluate cog-
nitive and vegetative symptoms of depression (range =
610 Rothbaum, Astin, and Marsteller
The Dissociative Experiences Scale-II (DES-II;
Bernstein & Putnam, 1986; Carlson & Putnam, 1993)
is a 28-item self-report scale used to quantify the fre-
quency and intensity of a wide range of experiences that
are indicative of absorption, dissociation, derealization,
amnesia, and depersonalization. In this newer version of
the original DES, participants indicate the percentage of
time they experience each item by circling a number from
0 to 100 (by 10s).
The State-Trait Anxiety Inventory (STAI: Spiel-
berger, Gorsuch, & Lushene, 1970) is a 40-item widely
used measure with two scales designed to assess state
anxiety and trait anxiety.
Integrity Measures
To insure treatment adherence and competence, two
session tapes from 50% of participants (25% of treatment
session tapes) receiving either EMDR or PE were indepen-
dently rated for treatment integrity. Dr. Francine Shapiro
designated an EMDR expert to make these ratings for
EMDR sessions, and Dr. Edna Foa designated a PE expert
from her lab to make these ratings for PE sessions. For PE,
the treatment integrity coding system was adapted from
one originally developed by Nishith and Resick (1994) in
a comparison study of PE and Cognitive Processing Ther-
apy which followed recommendations made by Waltz,
Addis, Koerner, and Jacobson (1993). It was modified for
this study in cooperation with Dr. Foa. A similar coding
system was developed for EMDR in cooperation with the
EMDR rater.
EMDR sessions were rated as 92.09% adherent for
essential and unique items while PE sessions were rated
90.46% adherent for items considered essential to each
protocol. Using a scale from 1 to 7, mean EMDR therapist
skill was rated 6.04 (SD = 0.58) or very good for essential
and unique items. Mean PE therapist skill was rated 5.80
(SD = 0.66) or very good for essential and unique items.
Participants were assigned to one of three doctoral-
level psychologists who were trained in both therapies.
To equate the two treatments, the EMDR protocol was
modified to match the standard nine-session PE protocol
used in previous studies. Thus, both treatments were de-
livered in nine 90-min, twice weekly sessions. In both,
the first two sessions consisted of information gathering,
education about trauma effects, a rationale for the par-
ticular treatment, and treatment preparation. Sessions 3
to 9 consisted of administration of PE or EMDR. Assess-
ments were conducted at pretreatment, posttreatment, and
follow-up of 6 and 12 months’ posttreatment. All assess-
ments were conducted by IAs who were kept blind to the
treatment condition.
The EMDR Technique. The EMDR manual dis-
tributed at the training workshops was the treatment man-
ual for this condition. EMDR involves having the pa-
tient imagine a scene that represents the worst part of the
trauma, focusing on t he sensations of distress in her body,
and rehearsing negative thoughts that match the picture.
The patient simultaneously follows the therapist’s fingers
moving back and forth approximately 18 in. in front of her,
a minimum of 20 times each repetition. Distress ratings
scale (SUDs). Once the distress about this scene from the
memory drops to 0 or 1, the patient is asked to track the
therapist’s finger while rehearsing a new, preferred belief,
repeating this sequence until the new statement feels true
to the patient. Cognitive work is accomplished through
the use of cognitive interweaves.
The PE Treatment. The PE treatment used in this
study is the same as that used in previous studies (Foa
& Rothbaum, 1998; Foa, Rothbaum, Riggs, Murdock, &
Walsh, 1991). A hierarchy of avoided situations is con-
structed for in vivo exposure homework. The next seven
sessions are devoted to reliving the rape scene in imag-
ination. Patients are instructed to try to imagine the as-
sault scene as vividly as possible and describe it aloud
in the present tense. Anxiety l evels (SUDs = 0–100) are
monitored every 5 min during exposure. Patients are en-
couraged to describe the rape in its entirety repeating it
several times for 45 to 60 min per session. Following ex-
posure, the patient’s reaction to the exposure situation is
discussed, and a homework assignment consonant with
that day’s exposure is assigned. The patient’s narratives
are tape-recorded, and they are instructed to listen to the
tapes at home at least once daily.
WAIT. The WAIT participants met with the IA for
their pretreatment assessment. They were scheduled for
the next session 4 to 5 weeks later (i.e., the amount of
time required for a course of treatment). WAIT partic-
ipants were randomly assigned to either PE or EMDR
and were provided with treatment free of charge after t he
posttreatment assessment.
Statistical Methods
We used multitrait, multimeasure multivariate re-
peated measure models for analysis to maximize power,
PE Versus EMDR for PTSD 611
which is essential to evaluate our second hypothesis. To
accomplish this, outcome measures were clustered into
two groups to be entered simultaneously into multivariate
repeated measures analyses. The first cluster consisted of
the PTSD frequency and intensity total-symptom scores
as measured by the CAPS, the PSS, and the IES. The
second cluster consisted of PTSD intrusion, avoidance,
and hyperarousal symptoms as measured by the CAPS,
the PSS, and the IES. To reduce scale effects, after con-
firming distribution assumptions, scores for each measure
were standardized to a mean of 0 and an SD of 1 for the
entire sample.
Hypotheses were tested sequentially. Pretreatment
to posttreatment change was analyzed first, with planned
contrasts of PE + EMDR versus WAIT and PE ver-
sus EMDR constructed from the 2 df for treatment.
Next, changes from pretreatment to posttreatment to
6-month follow-up were analyzed, using planned con-
trasts constructed to compare change from pretreatment
to 6 months and from posttreatment to 6 months. Co-
hen’s (1988, p. 471) effect sizes (ES) for multivariate
tests were obtained from the Wilks’s Lambda statistic us-
ing Cohen’s eq. 10.2.2 (1988, p. 473). For this expression
of ES, .35 is considered large, .15 is moderate, and .02 is
Categorical data were analyzed using log-linear
models containing the same planned contrasts defined
earlier at the end of treatment. When there were only
two treatment groups to consider, Fisher’s exact test was
Study Sample
Of the 74 women enrolled in the study, 1 dropped
out during the assessment phase, 1 was terminated and
referred during treatment for not meeting treatment cri-
teria, 12 dropped out during treatment, and 60 women
(83.3%) completed the protocol. The dropout rate across
the three groups was not significantly different, PE: 13.0%
(n = 3, 2 before MID); EMDR: 20.0% (n = 5, 4 before
MID); and WAIT: 16.7% (n = 4). Of the 40 active-
treatment participants who completed treatment, 10.0%
of PE subjects (n = 2) and 5.0% of EMDR subjects
(n = 1) were not interviewed at the 6-month follow-up.
Because only 2 of 14 participants who did not complete
the study (1 in each of the active treatments) provide
data other than baseline, intent-to-treat analyses provide
no consequentially different results and are not included
In the completer sample of 60 women, mean par-
ticipant age was 33.8 years (SD = 11.0). The majority
of participants were Caucasian (68.3%), never married
or divorced (73.3%), did not have children (68.3%), and
were employed full-time, part-time, or were full-time stu-
dents (76.7%). Almost half of the sample had earned at
least a college degree or more (48.3%), but only one third
(30.0%) had a household income over $40,000. There
were no significant differences among the three treatment
conditions for any of these variables. As assessed by the
SCID, 35% (n = 21) of participants had only a PTSD di-
agnosis, 40% (n = 24) had one comorbid diagnosis, and
25% (n = 15) had two or more diagnoses in addition to
PTSD. Fisher’s Exact Test for association of 0, 1, or 2+
comorbid diagnoses with treatment assignment was not
significant (p = .873).
Index assault experiences lasted an average of 87.98
min (SD = 144.63) and were perpetrated by one to
three assailants, with the majority (90%) perpetrated by
one assailant. Most assaults occurred in the residence of
the victim (28.3%) or the perpetrator (21.7%), but also
were perpetrated in other residences (6.7%), abandoned
buildings (3.3%), vehicles (11.7%), outdoors (18.3%), or
other settings (11.7%). The majority of assaults (43.4%)
were perpetrated by friends, relatives, dates, and signifi-
cant others; 33.3% by strangers; and 23.3% by acquain-
tances. Including the index assault, participants experi-
enced a mean of 6.0 traumas (SD = 4.1) prior to study
entry. Mean time since assault in months varied greatly
(EMDR: 145.9, SD = 146.8; PE: 120.9, SD = 94.1;
WAIT: 162.9, SD = 136.9). There were no significant
differences among the three treatment conditions on any
of these variables.
Baseline Measures
Least-square means of the total scores for each of
the three measures of PTSD symptoms are shown in
Fig. 1 for each time point. Despite randomization, compar-
isons between the three groups at pretreatment revealed
significant differences on some measures. As assessed
by the CAPS, participants in the EMDR condition ex-
hibited significantly higher overall PTSD symptoms,
612 Rothbaum, Astin, and Marsteller
Fig. 1. Total Scores on the CAPS, IES, and PSS at pretreatment,
posttreatment, and 6-month follow-up (least-square means ± 95%
confidence interval). Note. CAPS = Clinician Administered PTSD
Scale; IES = The Impact of Events Scale; PSS = PTSD symptom Scale.
PRE = pretreatment; POST = posttreatment; 6 mo FU = 6-month
F(1, 57) = 9.2, p<.01, due apparently to levels of
avoidance symptoms, F (1, 57) = 13.7, p<.001, than
did PE participants. No differences between groups
emerged on self-report measures of PTSD (PSS and
IES-R) except that EMDR participants reported higher
levels of intrusive symptoms on the PSS than did PE
participants, F (2, 57) = 5.0,p < .05. The EMDR group
also exhibited significantly higher levels of depression
(BDI), F (1, 57) = 11.3,p < .001, dissociation (DES-
II), F (1, 57) = 7.4,p = .01, and trait anxiety (STAI-T),
F (1, 57) = 5.0,p = .05, but not state anxiety (STAI-S),
F (1, 57) = 3.7,p = .059, than the PE group. The re-
peated measures analyses used to compare treatments
analyze the change in scale scores from baseline, thus
controlling for these pretreatment differences in symptom
Change From Pretreatment to Posttreatment
At posttreatment, 5% of PE participants (n = 1),
25% of EMDR participants (n = 5), and 90% of WAIT
participants (n = 18) still met criteria for a diagnosis
of PTSD. Log-linear models partitioned in the same
way as the continuous models found that significantly
fewer participants in PE and EMDR were PTSD pos-
itive at posttreatment than WAIT participants, χ
(n =
1) = 20.10,p < .001; however, the difference between
active treatments (PE vs. EMDR) in PTSD diagnostic sta-
tus was not significant, χ
(n = 1) = 2.58,p = .108. At
6-months follow-up, 5.6% of PE participants (n = 1) and
26.3% of EMDR participants (n = 5) continued to receive
a diagnosis of PTSD, Fisher’s exact p = .185.
The repeated measures test of composite total score
improvement from pre- to posttreatment was significant,
F (1, 57) = 108.8,p < .001 (ES = 1.91), as were Time ×
Treatment interaction effects, F (2, 57) = 16.0,p < .001
(ES = 0.56). Planned contrasts showed that active-
treatment participants (PE or EMDR) improved sig-
nificantly more than WAIT participants, F (1, 57) =
31.7,p < .001 (ES = 0.56), but the change from pre-
treatment to posttreatment did not differ significantly be-
tween PE and EMDR, F (1, 57) = 0.3,p = .608 (ES =
.005). Results from the PTSD symptom cluster MANOVA
and analyses of individual measures and symptoms were
similar. Overall improvement (time effect) was signifi-
cant from pre- to posttreatment, F (1, 57) = 116.6,p <
.001 (ES = 2.05), and there was a significant treat-
ment effect for improvement, F (2, 57) = 16.3,p < .001
(ES = 0.57). Active-treatment participants improved
significantly more than WAIT participants, F (1, 57) =
32.1,p < .001 (ES = 0.56). There were no significant
differences in symptom cluster changes from pretreat-
ment to posttreatment between PE and EMDR groups,
F (1, 57) = 0.5,p < .5(ES= 0.01).
End-State Functioning
Composite measures were used to examine good
end-state functioning. Good end-state functioning was de-
fined as a combined criteria on three measures: 50% or
more decrease on the CAPS from pretreatment, a score
PE Versus EMDR for PTSD 613
of 10 or less on the BDI, and a score of 40 or less on
the STAI-S. At posttreatment, 70% (n = 14) of PE par-
ticipants, 50% of EMDR participants (n = 10), and none
of the WAIT participants met these criteria. Differences
were not significant between PE versus EMDR groups, but
were significant between Treatment versus WAIT groups,
Fisher’s exact p = .001. At 6-months follow-up, 78%
(n = 14) of PE participants and 35.3% (n = 6) of EMDR
participants met criteria for good end-state functioning.
Significantly more PE participants met this criteria than
EMDR participants at 6-months follow-up, Fisher’s Exact
Test p = .017. These are measures of status at a point in
time and are therefore sensitive to differences in initial
6-Month Follow-Up
There was no WAIT at the 6-month follow-up as all
WAIT participants were offered treatment after comple-
tion of the WAIT period. PTSD Total scores improved sig-
nificantly from pretreatment through 6-month follow-up,
F (2, 33) = 68.3,p < .001 (ES = 4.14), with no signif-
icant difference between the two treatments, F (2, 33) =
1.8,ns (ES = 0.11). PTSD symptom clusters also im-
proved significantly from pretreatment through 6-month
follow-up, F (2, 33) = 74.8,p < .001 (ES = 4.40), with
no significant differences between treatments, F (2, 33) =
0.9,ns (ES = 0.06). Planned multivariate contrasts com-
paring change from pretreatment to 6 months and post-
treatment to 6 months did not show significant differences
Table 2. SD Other Symptom Self-Report Treatment = Outcome Differences × Treatment Condition
Treatment condition
Measure Assessment EMDR PE WAIT
BDI PRE 25.95 (7.11) 16.70 (8.18) 24.05 (10.50)
POST 10.70 (11.45) 4.65 (4.99) 22.20 (10.55)
6 MO 10.53 (10.92) 4.44 (5.07) -
DES-II PRE 18.68 (12.67) 10.13 (5.45) 12.53 (10.18)
POST 8.12 (7.98) 4.84 (4.65) 12.36 (8.51)
6 MO 8.91 (9.10) 3.51 (2.61) -
STAI-State PRE 51.10 (11.05) 43.33 (12.59) 46.58 (13.48)
POST 32.60 (11.62) 30.00 (10.44) 49.00 (13.73)
6 MO 38.89 (14.54) 29.19 (8.79) -
STAI-Trait PRE 56.80 (10.95) 48.72 (8.62) 53.42 (13.07)
POST 41.10 (14.48) 35.56 (9.88) 53.95 (13.01)
6 MO 41.44 (13.26) 34.19 (7.52) -
Note. EMDR = Eye Movement Desensitization and Reprocessing; PE = Prolonged Exposure; WAIT
= Waitlist control; PRE = pretreatment; POST = posttreatment; 6 MO = 6-month follow-up; BDI =
Beck Depression Inventory; DES-II = Dissociative Experiences Scale-II; STAI = State-Trait Anxiety
Inventory; At Pretreatment, n = 60 for BDI and DES-II and n =58 for STAI-State and STAI-Trait;
At Posttreatment, n = 60 for BDI, n = 59 for DES-II and STAI-State, and n = 58 for STAI-Trait; At
6-month follow-up, n =37 for BDI and DES-II, and n = 36 for STAI-State and STAI-Trait.
between the treatments for either interval for Total scores
or for symptom composites.
Secondary Measures
Overall, depression scores (Table 2) decreased sig-
nificantly from pretreatment to posttreatment, F (1, 57) =
67.1,p < .001, and participants in the active treatments
improved significantly more than did WAIT partici-
pants, F (2, 57) = 22.0,p < .001. Improvement of PE
and EMDR participants did not differ from pretreatment to
posttreatment, F (1, 57) = 1.2,ns, or through 6 months,
F (2, 34) = 0.6,ns. Given the significant initial differ-
ences and similar improvements, participants assigned to
PE and EMDR differed in the BDI scores at each time
Dissociative symptoms (Table 2) also significantly
decreased from pre- to posttreatment, F (1, 56) = 25.6,
p<.001, and symptoms decreased significantly more
in active versus WAIT conditions, F (1, 56) = 12.1,p <
.001. Scores improved significantly more from pretreat-
ment to posttreatment in the EMDR group F (1, 56) =
4.1,p < .05. This marginal difference became nonsignif-
icant when 6-month scores were considered, F (2, 33) =
614 Rothbaum, Astin, and Marsteller
For state anxiety (Table 2), pretreatment to post-
treatment scores decreased significantly, F (1, 54) =
35.5,p < .001, and participants in the active treatments
improved significantly more than those in the WAIT
condition, F (1, 36) = 27.7,p < .001. No significant dif-
ferences in improvement of state anxiety were found
between PE and EMDR from pretreatment to either
posttreatment or 6 months. Similarly, trait anxiety de-
creased significantly from pretreatment to posttreatment,
F (1, 54) = 36.2,p < .001, and participants in the active
treatments improved significantly more than did WAIT
participants, F (1, 54) = 20.2,p < .001. In general, pre-
treatment differences between groups in these secondary
measures were retained at all time points.
In summary, both PE and EMDR equally led to clin-
ically and statistically significant improvements immedi-
ately following treatment compared to the waitlist control
condition. At posttreatment, 95% of PE participants and
75% of EMDR participants no longer met criteria for
PTSD, which was not significantly different. For the most
part, gains were maintained at 6-months posttreatment;
however, at that 6-month follow-up, PE participants ev-
idenced higher end-state functioning (78 vs. 35%) than
did EMDR participants, but not a significantly lower r ate
of PTSD diagnosis, and again, there were pretreatment
differences on some measures that could have influenced
this composite measure.
The particular methodology of the current study and
implications should be noted. In this study, the goal was
to compare good PE to good EMDR, so it should be clear
that the data presented here cannot answer the question
regarding the necessity of the eye movements. It also at-
tempted to equate PE and EMDR as much as possible, so
EMDR was expanded to nine sessions. Sessions 1 and 2
were nearly identical for both treatments and followed the
typical PE format of gathering information and providing
an explanation of PTSD and rationale for treatment.
The results of this study differ from the results of
two other recent studies of EMDR compared to CBT
(Devilly & Spence, 1999; Taylor et al., 2003) in that
treatments differed in improvement at posttreatment or
long-term follow-up in those studies. This view is com-
plicated by comparatively large differences in categorical
end-state functioning in this study, but this measure is
strongly influenced by scores of anxiety and depression
scales that (a) differed at baseline between PE and EMDR
participants and (b) are not direct targets of the therapies
It is our contention that EMDR and PE are both
exposure techniques, assisting the patient in confronting
her assault memory and repeating this exposure until she
can remember it with low or no anxiety. Both depend
on imaginal exposure (i.e., having the patient obtain a
picture of the traumatic image in her mind’s eye). Both
are presumed to aid emotional processing of the trauma,
and both aim to achieve cognitive modifications via the
treatment process.
The treatment techniques diverge in the administra-
tion and between-session instructions. As noted earlier,
the exact role of the eye movements in the therapeutic
process is unclear, but is thought by these authors to pos-
sibly act as a distracter which complicates the task and
possibly undermines avoidance tendencies that may al-
low the patient to tolerate the image. In contrast, EMDR
proponents (e.g., Lipke, 2003) have pointed to nonclinical
studies which have found that eye movements decrease the
vividness of memories and related emotions and increases
in memory access. For example, Kavanagh, Freese,
Andrade, and May (2001) proposed that eye movements
interrupt working memory, which decreases the vividness
of memories and thus leads to decreased affect. They ar-
gued eye movements may titrate the exposure t o memories
and related emotions.
The blank it out aspect of EMDR also is a difference
between the two treatments; after t hinking about a partic-
ular aspect of the trauma during the eye-movement phase,
the participant is instructed to let her mind go blank for a
moment before continuing. PE allows the dose of expo-
sure to be titrated up by allowing the patient to gloss over
more traumatic details in the first few sessions, gradually
requesting more and more until focusing just on the hot
spots repeatedly. In EMDR, the patient gets breaks from
the image every few minutes when requested to blank it
out. In PE, the patient usually engages in more repetitions
of the same traumatic scene than in EMDR. For exam-
ple, in PE, if it takes 15 min for the patient to recount
what happened, it will usually be reviewed four times in a
session. Additionally, in PE, patients are assigned home-
work to listen to the audiotaped accounts of their retelling
and to practice in vivo exposure whereas no homework
is assigned in EMDR other than to notice any changes
in thoughts and feelings between sessions. An interest-
ing potential clinical implication is that EMDR seemed
to do equally well in the main despite less exposure and
no homework. It will be important for future research to
explore these issues.
The current study has many important strengths. It
was methodologically rigorous and meets all seven of the
PE Versus EMDR for PTSD 615
gold standards for a treatment-outcome study including
clearly defined symptoms; reliable and valid measures;
use of independent evaluators; trained assessors; manu-
alized, replicable, specific treatment programs; treatment
adherence; and unbiased assignment to treatment (Foa &
Meadows, 1997). Table 1 compares EMDR studies on
these seven gold standards and compares two active and
intensive, but brief, treatments to a control condition. Fur-
thermore, although some exclusionary criteria were nec-
essary, this study attempted to keep these to a minimum
to treat real-world PTSD sufferers so that results could be
generalized to other PTSD clients.
This research was supported by NIMH Grant 1 R01
MH56351-01A1 awarded to the first author.
American Psychiatric Association. (1994). Diagnostic and statistical
manual of mental disorders (4rth ed.). Washington, DC: Author.
Barrowcliff, A.L., Gray, N.S., MacCulloch, S., Freeman, T.C.A., &
MacCulloch, M.J. (2003). Horizontal rhythmical eye movements
consistently diminish the arousal provoked by auditory stimuli.
British Journal of Clinical Psychology, 42, 289–302.
Beck, A.T., Steer, R.A., & Garbin, M.G. (1988). Psychometric properties
of the Beck Depression Inventory: Twenty-five years of evaluation.
Clinical Psychology Review, 8, 77–100.
Beck, A.T., Ward, C.H., Mendelsohn, M., Mock, J., & Erbaugh, J.
(1961). An inventory for measuring depression. Archives of Gen-
eral Psychiatry, 4, 561–571.
Bernstein, E.M., & Putnam, F.W. (1986). Development, reliability, and
validity of a dissociation scale. Journal of Nervous and Mental
Disease, 174, 727–735.
Blake, D.D., Weathers, F.W., Nagy, L.M., Kaloupek, D.G., Gusman,
F.D., Charney, D.S., et al. (1995). The development of a clinician-
administered PTSD scale. Journal of Traumatic Stress, 8, 75–90.
Blake, D.D., Weathers, F.W., Nagy, L.M., Kaloupek, D.G., Klauminzer,
G., Charney, D.S., et al. (1990). A clinician rating scale for assessing
current and lifetime PTSD: The CAPS-1. The Behavior Therapist,
18, 187–188.
Carlson, E.B., & Putnam, F.W. (1993). An update on the Dissociative
Experiences Scale. Dissociation: Progress in the Dissociative Dis-
orders, 6, 16–27.
Chemtob, C.M., Tolin, D.F., van der Kolk, B.A., & Pitman, R.K. (2000).
Guidelines for treatment of PTSD: Eye movement desensitization
and reprocessing. Journal of Traumatic Stress, 13, 569–570.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences.
Hillsdale, NJ: Academic Press.
Devilly, G.J. (2001). The influence of distraction during exposure and
researcher allegiance during outcome trials. Behvavior Therapist,
24, 18–21.
Devilly, G.J., & Spence, S.H. (1999). The relative efficacy and treatment
distress of EMDR and a cognitive behavioral treatment protocol
in the amelioration of post traumatic stress disorder. Journal of
Anxiety Disorders, 13, 131–157.
First, M., Gibbons, M., Spitzer, R.L., & Williams, J.B.W. (1996). Struc-
tured Clinical Interview for DSM-IV (SCID).New York: New York
State Psychiatric Institute, Biometrics Research Department.
Foa, E.B., Friedman, M., & Keane, T. (2000). Effective treatments for
posttraumatic stress disorder: Practice guidelines from the Interna-
tional Society for Traumatic Stress Studies. New York: Guilford
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 post-
traumatic stress disorder: A critical review. Annual Review of Psy-
chology, 48, 449–480.
Foa, E.B., Riggs, D.S., Dancu, C.V., & Rothbaum, B.O. (1993). Relia-
bility and validity of a brief instrument for assessing post-traumatic
stress disorder. Journal of Traumatic Stress, 6, 459–473.
Foa, E.B., & Rothbaum, B.O. (1998). Treating the trauma of rape:
A cognitive-behavioral therapy for PTSD. New York: Guilford
Foa, E., Rothbaum, B.O., Riggs, D., Murdock, T., & Walsh, W. (1991).
Treatment of posttraumatic stress disorder in rape victims: A
comparison between cognitive-behavioral procedures and coun-
seling. Journal of Consulting and Clinical Psychology, 59, 715–
Foa, E.B., Steketee, G., & Rothbaum, B. (1989). Behavioral/cognitive
conceptualizations of post-traumatic stress disorder. Behavior Ther-
apy, 20, 155–176.
Goodman, L.A., Corcoran, C., Turner, K., Yuan, N., & Green, B.L.
(1998). Assessing traumatic event exposure: General issues and
preliminary findings for The Stressful Life Events Screening Ques-
tionnaire. Journal of Traumatic Stress, 11, 521–542.
Horowitz, M.J., Wilner, N., & Alvarez, W. (1979). Impact of event
scale: A measure of subjective distress. Psychosomatic Medicine,
41, 207–218
Ironson, G., Freund, B., Strauss, J.L., & Williams, J. (2002). Comparison
for two treatments for traumatic stress: A community-based study
of EMDR and prolonged exposure. Journal of Clinical Psychology,
58, 113–128.
Kavanagh, D.J., Freese, S., Andrade, J., & May, J. (2001). Effects of
visuospatial tasks of desensitization to emotion memories. British
Journal of Clinical Psychology, 40, 267–280.
Kilpatrick, D.G., & Calhoun, K.S. (1988). Early behavioral treatment
for rape trauma: Efficacy or artifact? Behavior Therapy, 19, 421–
Lee, C., Gavriel, H., Drummond, P., Richards, J., & Greenwald, R.
(2002). Treatment of PTSD: Stress inoculation training with pro-
longed exposure compared to EMDR. Journal of Clinical Psychol-
ogy, 58, 1071–1089.
Lipke, H. (2003). Comment on Hembree and Foa (2003) and EMDR.
Journal of Traumatic Stress, 16, 573–574.
Lohr, J.M., Tolin, D.F., & Lilienfeld, S.O. (1998). Efficacy of eye move-
ment desensitization and reprocessing: Implications for behavior
therapy. Behavior Therapy, 29, 123–156.
Nishith, P., & Resick, P.A. (1994). Adherence and competence rating
scale for prolonged exposure treatment. Unpublished rating scale.
University of Missouri-St. Louis.
Resnick, H.S., Kilpatrick, D.G., Dansky, B.S., Saunders, B.E., & Best,
C.L. (1993). Prevalence of civilian trauma and post-traumatic stress
disorder in a representative national sample of women. Journal of
Consulting and Clinical Psychology, 61, 984–991.
Rogers, S., & Silver, S.M. (2002). Is EMDR an exposure therapy? A
review of trauma protocols. Journal of Clinical Psychology, 58,
Rogers, S., Silver, S.M., Goss, J., Obenchain, J., Willis, A., & Whitney,
R.L. (1999). A single session, group study of exposure and eye
movement desensitization and reprocessing in treating posttrau-
matic stress disorder among Vietnam War veterans: Preliminary
data. Journal of Anxiety Disorders, 13, 119–130.
Rothbaum, B.O. (1997). A controlled study of eye movement desensi-
tization and reprocessing in the treatment of posttraumatic stress
disordered sexual assault victims. Bulletin of the Menninger Clinic,
61, 1–18.
Rothbaum, B.O., Dancu, C.V., Riggs, D., & Foa, E.B. (1990, September).
The PTSD Symptom Scale. Presented at the European Association
of Behavior Therapy Convention, Paris.
Rothbaum, B.O., & Foa, E.B. (1989). Subtypes of PTSD and duration
of symptoms. Position paper prepared for the DSM-IV Anxiety
Disorders Workgroup, PTSD Subgroup.
616 Rothbaum, Astin, and Marsteller
Rothbaum B.O., Foa, E.B., Riggs, D., Murdock, T., & Walsh, W.
(1992). A prospective examination of post-traumatic stress dis-
order in rape victims. Journal of Traumatic Stress, 5, 455–
Rothbaum, B.O., Meadows, E.A., Resick, P., & Foy, D.W. (2000).
Cognitive-behavioral treatment position paper summary for the
ISTSS Treatment Guidelines Committee. Journal of Traumatic
Stress, 13, 558–563.
Simon, M.J. (2000), A comparison between EMDR and exposure for
treating PTSD: A single-subject analysis. Behavior Therapist, 23,
Spielberger, C.D., Gorsuch, R.L., & Lushene, R.E. (1970). Manual for
the State-Trait Anxiety Inventory (self-evaluation questionnaire).
Palo Alto, CA: Consulting Psychologists Press.
Taylor, S., Thordarson, D.S., & Maxfield, L., Fedoroff, I.C., Lovell, K.,
& Ogrodniczuk, J. (2003). C omparative efficacy, speed, and adverse
effects of three PTSD treatments: Exposure therapy, EMDR, and
relaxation training. Journal of Consulting and Clinical Psychology,
71, 330–338.
Waltz, J., Addis, M.E., Koerner, K., & Jacobson, N.S. (1993). Testing
the integrity of a psychotherapy protocol: Assessment of adherence
and competence. Journal of Consulting and Clinical Psychology,
61, 620–630.
Weathers, F.W., Blake, D.D., Krinsley, K.E., Haddad, W., Huska, J.,
& Keane, T.M. (1992). The clinician administered PTSD scale:
Reliability and construct validity. Paper presented at the 26th annual
meeting of the Association for the Advancement of Behavioral
Therapy, Boston.
Weiss, D.S., & Marmar, C.R. (1997). The Impact of Event Scale-
Revised. In J.P. Wilson & T.M. Keane (Eds.), Assessing psy-
chological trauma and PTSD (pp. 399–411). New York: Guilford
    • "It also resulted in reduced maternal stress and avoidance symptoms. Sexual violence Eye movement Desensitization and Reprocessing (Rothbaum, Astin, & Marsteller, 2005) Cognitive processing Therapy (Resick, Williams, Suvak, Monson, & Gradus, 2012) Individual weekly to twice weekly sessions (6– 13.5 h of treatment) "
    Article · Aug 2016
    • "The participants in EMDR RCTs have often experienced a mixture of trauma types and frequently included childhood sexual abuse and adult sexual assault survivors in the samples (Edmond & Lawrence, 2015 ). In addition, several studies have tested the effectiveness of EMDR in samples composed entirely of female childhood sexual abuse or sexual assault survivors (Edmond, Rubin, & Wambach, 1999; Edmond & Rubin, 2004; Jaberghaderi, Greenwald, Rubin, Dolatabadim, & Zand, 2004; Rothbaum, 1997; Rothbaum, Astin, & Marsteller, 2005; Scheck, Schaeffer, & Gillete, 1998). Collectively, the existing empirical data on EMDR provide ample evidence of its effectiveness in treating PTSD, depression, and other trauma symptoms in child, adolescent, and adult sexual abuse/assault survivors. "
    [Show abstract] [Hide abstract] ABSTRACT: Sexual violence is pervasive and generates significant trauma symptoms that can last a lifetime for survivors. Rape crisis centers provide critically important services for survivors of child sexual abuse and adult sexual assault, including individual and group counseling. Eye movement desensitization and reprocessing (EMDR) has been found to be an effective treatment for a wide array of trauma symptoms in both children and adults. This study sought to determine the extent to which rape crisis centers use EMDR therapy, practitioners' perceptions of EMDR, and the provider characteristics that might support or hinder implementation of EMDR in this setting. A statewide web-based survey generated responses from 76 counselors working within 47 rape crisis centers. Results indicate that there is a low-use rate of EMDR (8%) in this setting, perceptions of EMDR were predominately marked by uncertainty, reflecting a lack of familiarity, but there is strong interest in receiving training. The desire for training is complicated by the range of education levels of counseling staff in rape crisis centers with only 54% holding advanced degrees. There is an opportunity and need to build capacity for the implementation of EMDR in this vital service sector, but there are also significant challenges that will need to be addressed.
    Full-text · Article · Feb 2016
    • "In a sample of childhood sexual abuse victims, after exposure therapy, self-reported symptoms of dissociation (Dissociative Experiences Scale [DES] [112, 113]) decreased, though this decline was not significant [60]. In rape victims, self-reported symptoms of dissociation (DES) decreased significantly after PE [44]. In victims of mixed trauma, symptoms of numbing (assessed via the Clinician- Administered PTSD Scale [CAPS] numbing subscale) significantly declined after PE [51]. "
    [Show abstract] [Hide abstract] ABSTRACT: Prolonged exposure (PE) is an effective psychological treatment for patients who suffer from PTSD. The majority of PTSD patients have comorbid psychiatric disorders, and some clinicians are hesitant to use PE with comorbid patients because they believe that comorbid conditions may worsen during PE. In this article, we reviewed the evidence for this question: what are the effects of PE on comorbid symptoms and associated symptomatic features? We reviewed findings from 18 randomized controlled trials of PE that assessed the most common comorbid conditions (major depression, anxiety disorders, substance use disorders, personality disorders, and psychotic disorders) and additional symptomatic features (suicidality, dissociation, negative cognitions, negative emotions, and general health and work/social functioning). Although systematic research is not available for all comorbid populations, the existing research indicates that comorbid disorders and additional symptomatic features either decline along with the PTSD symptoms or do not change as a result of PE. Therefore, among the populations that have been studied to date, there is no empirical basis for excluding PTSD patients from PE due to fear of increases in comorbid conditions or additional symptomatic features. Limitations of the existing research and recommendations for future research are also discussed.
    Full-text · Article · Mar 2015
Show more