2177

ORIGINAL RESEARCH—PSYCHOLOGY

Effects of Expressive Writing on Sexual Dysfunction, Depression,

and PTSD in Women with a History of Childhood Sexual Abuse:
Results from a Randomized Clinical Trial

Cindy M. Meston, PhD, Tierney A. Lorenz, MA, and Kyle R. Stephenson, MA

Department of Psychology, University of Texas at Austin, Austin, TX, USA

DOI: 10.1111/jsm.12247

ABSTRACT

Introduction. Women with a history of childhood sexual abuse (CSA) have high rates of depression, posttraumatic
stress disorder, and sexual problems in adulthood.
Aim. We tested an expressive writing-based intervention for its effects on psychopathology, sexual function,
satisfaction, and distress in women who have a history of CSA.
Methods. Seventy women with CSA histories completed five 30-minute sessions of expressive writing, either with a
trauma focus or a sexual schema focus.
Main Outcome Measures. Validated self-report measures of psychopathology and sexual function were conducted at
posttreatment: 2 weeks, 1 month, and 6 months.
Results. Women in both writing interventions exhibited improved symptoms of depression and posttraumatic stress
disorder (PTSD). Women who were instructed to write about the impact of the abuse on their sexual schema were
significantly more likely to recover from sexual dysfunction.
Conclusions. Expressive writing may improve depressive and PTSD symptoms in women with CSA histories. Sexual
schema-focused expressive writing in particular appears to improve sexual problems, especially for depressed women
with CSA histories. Both treatments are accessible, cost-effective, and acceptable to patients. Meston CM, Lorenz
TA, and Stephenson KR. Effects of expressive writing on sexual dysfunction, depression, and PTSD in
women with a history of childhood sexual abuse: Results from a randomized clinical trial. J Sex Med
2013;10:2177–2189.
Key Words. Childhood Sexual Abuse; Sexual Satisfaction; Sexual Function; Expressive Writing; Depression; Sexual
Dysfunction

Introduction cal samples [2,3]. Survivors of CSA also report

significantly lower sexual satisfaction and higher
I t is not surprising that survivors of childhood
sexual abuse (CSA) report significant problems
in adulthood with intimate relationships, particu-
sexual distress [4]. Beyond the impact on quality
of life, sexual problems contribute to relationship
larly sexual relationships [1].1 Women who report distress [1] and have been implicated in the high
histories of CSA have significantly higher rates of divorce rates seen in CSA survivors [5].
sexual dysfunction than either community or clini- What is surprising is that, despite the preva-
lence of CSA (over 66,000 reports of CSA were
1 substantiated in 2009 [6], with many more cases
Although both males and females may experience CSA,
the vast majority of cases involves girls and, as such, the that go unreported [7]), there are few treatments
majority of research and clinical focus has been on female for sexual dysfunction that have been empirically
survivors of CSA. validated for adult survivors. Treatments for sexual

© 2013 International Society for Sexual Medicine J Sex Med 2013;10:2177–2189

2178 Meston et al.

problems that were developed in women without control but showed no difference in sexual con-
abuse histories have shown inconsistent results in cerns. From these results, it is clear that sexual
women with abuse histories. For example, tradi- problems are often distinct from psychopathology
tional sex therapy techniques such as sensate focus and may require separate clinical focus.
are overwhelming for many women with CSA his- To that end, there have been only three peer-
tories [8]. Similarly, pharmacological treatments reviewed reports on psychotherapy with adult
that improve sexual response in women without survivors of CSA that have demonstrated improve-
CSA histories, such as sildenafil, do not improve ments on a validated measure of sexual function.
and, in some cases, may even worsen sexual prob- Hazzard, Rogers, and Angert [18] found that for
lems in women with CSA histories [9]. This may the 102 participants who completed a full year of
be because for survivors of CSA, the sexual weekly process-oriented group therapy sessions,
response can be a powerful reminder of the there was a significant decrease in sexual avoidance
trauma. and sexual dysfunction as per the Sexual Symptom
It is often suggested that psychopathology such Checklist; however, as this study did not include a
as depression or traumatic stress related to CSA control condition, it is difficult to know the spe-
should be resolved before addressing sexual prob- cific efficacy of the treatment. Hébert and Berg-
lems. For example, from a prominent workbook eron [19], on the other hand, did include a waitlist
for CSA-related therapy issues: “it is recom- control in their trial of semi-structured group
mended that therapists address more general therapy for CSA survivors. A total of 41 women
effects of sexual abuse, such as depression, anger, completed 15–17 weeks of 3-hour sessions that
self-blame, self-destructive behaviors, and trust included discussion, relaxation techniques, and art
concerns, before doing work on sexual problems” therapy. Treatment completers reported signifi-
[10]. Others have posited models in which psycho- cantly decreased sexual anxiety and discomfort
pathology mediates the relationship between CSA about their sexuality as measured by the Multidi-
survivorship and sexual problems [11,12], which in mensional Sexual Self-Concept Scale, whereas
turn implies resolution of psychopathology would participants in the control condition remained
lead to a resolution of sexual problems. Consider- stable. Treatment effects were maintained at a
ing that women with CSA histories are 1.3–2.2 3-month follow-up. While promising, the treat-
times more likely to report major depression or ment was not standardized and thus difficult to
other mood disorders and 2.1–2.6 times more replicate. Finally, Brotto, Basson, and Luria [20]
likely to report posttraumatic stress disorder conducted an uncontrolled trial of a manualized,
(PTSD) than women who have never been sexu- three-session psychoeducation intervention for
ally abused [13], this seems reasonable. sexual desire and arousal disorders in 15 women.
However, sexual problems are often present in In a post hoc analysis, they found that the eight
CSA survivors even in the absence of symptoms of women in the therapy group with a history of CSA
depression or traumatic stress [14,15] and may improved significantly in measures of sexual func-
persist even after resolution of psychopathology tion and sexual distress on the Female Sexual Dis-
through successful treatment [16]. For example, tress Scale and the Female Sexual Function Index,
Rieckert and Möller [15] conducted a trial of whereas the 17 women without such a history did
group rational emotive behavior therapy for 42 not show any significant change. With so few par-
women with CSA histories. After 10 weekly ses- ticipants and an exploratory post hoc analysis, the
sions, the participants in the treatment group authors cautioned that these results should be
moved from the severe depression range of the treated as preliminary. In short, while there is
Beck Depression Inventory to the normal range; some evidence that psychotherapy may improve
this improvement was maintained at the 8-week sexual problems in women with CSA histories,
follow-up. However, participants who received there is little direct evidence.
treatment did not differ from control participants One common theme in these few studies
on a validated measure of sexual function and sat- reporting improvements in sexual function in CSA
isfaction. Similarly, Classen et al. [17] studied the survivors is a focus on the impact of CSA on
effects of a trauma-focused group therapy protocol women’s thoughts, beliefs, and feelings related
in 166 CSA survivors. Intent-to-treat analyses to sexuality. In other words, these treatments
revealed that after 24 weekly sessions, participants attempted, directly or indirectly, to address nega-
in the treatment condition had significantly tive sexual schema in women with CSA histories.
reduced PTSD severity relative to a waitlist Sexual schemas are cognitive structures that give

J Sex Med 2013;10:2177–2189

CSA and Sex Function 2179

meaning and order to thoughts and feelings about freely as possible [34]. The assigned writing topic
oneself as sexual being and sexuality in general. is generally a stressful, emotional, or traumatic
Sexual schemas help guide representations of event [34], but writing about other topics such as
memories of sexual experiences, plan sexual behav- body image [36], adjustment to college [37], and
ior, and interpret responses to sexual stimuli [21]. sexual orientation [38] have also been explored.
The cornerstones of schema are beliefs, which Expressive writing has been demonstrated to
translate emotionally significant life events improve psychological and physical health in a
and social learning into organizing principles for wide variety of settings and populations. Several
information about the self, others, and the world meta-analyses have estimated effect sizes ranging
[22]. These principles operate at a preconscious from d = 0.15 [39] to d = 0.47 [40] for positive out-
level, orienting attention toward or away from comes ranging from reduced medical care usage to
sexual stimuli, and at a conscious level, directing posttraumatic growth.
sexual attitudes and behaviors [23]. Sexual beliefs, Specifically, expressive writing has been shown
and the corresponding schema, may be positive to improve both depression and PTSD [41].
(e.g., “I am a passionate woman”) or negative (“I Improvements due to writing have been linked to
am an unloving woman”). Positive sexual schemas increased cognitive processing, above and beyond
regarding the self, or sexual self-schema (SSS), are expression of emotions [42]. Writing leads to
associated with higher sexual satisfaction [24], increased use of insight and causation-related
whereas negative SSS are associated with greater words [43], which are associated with the construc-
sexual distress and dysfunction [25]. Negative SSS tion of a cognitive structure [44]. Writing about
may be a diathesis or vulnerability factor for sexual traumatic experiences has been shown to be as
dysfunction [26]. effective as cognitive therapy in improving
Women with CSA histories have significantly trauma-related beliefs, including beliefs about inti-
more negative SSS, which likely contribute to macy, in female survivors of interpersonal violence
sexual problems [27]. At an unconscious level, [45].
women with CSA histories are less likely to asso- To that end, expressive writing has been tested
ciate sexual stimuli with positive emotions than as a treatment for the psychological sequelae of
women without CSA histories [23,28–30]. At a intimate violence against women, although results
conscious level, women with CSA histories are have been mixed. One study investigating under-
more likely to endorse SSS that cast themselves as graduate women with a history of sexual assault
immoral or irresponsible [31] and less likely to found that trauma-focused writing was not supe-
endorse romantic or passionate SSS [27,32,33]. rior to writing about trivial topics in improving
Low endorsement of positive SSS has been found PTSD symptoms [46]. However, as this was not a
to mediate negative affect during sex for CSA sur- clinical population, symptoms of PTSD were low
vivors independently of depression and anxiety at baseline and thus had little room for further
symptoms [27]. improvement. Another study of adult survivors of
One powerful yet simple way to impact schema CSA found that expressive writing did not improve
is through writing. Constructing a written narra- depressive symptoms [47]. The authors noted that
tive about an emotional event helps to integrate while their paradigm involved sessions on con-
that experience into existing schema and high- secutive days, meta-analyses have shown that time
lights the meaningful aspects of the experience to between sessions moderates the benefits of expres-
help construct new schema [34]. Writing about sive writing [40], and thus there may not have been
traumatic events has been shown to help individu- sufficient time for participants to process the
als make meaning of these experiences, adopt less content of their writing between sessions. In con-
aversive appraisals of the event, and process the trast to these two studies, Koopman et al. [48]
experience in a larger context [35]. Such cognitive studied a writing treatment spaced out over several
processing through writing may help to improve weeks in a nonstudent sample of women with a
implicit attitudes toward the self, and reorganize history of intimate partner violence. There was a
self-schema [36]. significant interaction of treatment and level of
Of particular promise is expressive writing, a depression at intake, such that women who
structured writing paradigm in which people write entered the study with a high level of depression
for a specified amount of time, generally from gained the most benefit in the expressive writing
15–45 minutes, during which they are encouraged condition, whereas women with low levels of
to express their deepest thoughts and feelings as depression or women in the control condition did

J Sex Med 2013;10:2177–2189

2180 Meston et al.

not benefit. These results were promising and where they were given more information about the
indicated that for those women with significant study and screened for inclusion and exclusion cri-
psychological distress related to sexual abuse, teria (see below). Following determination of eli-
expressive writing may confer a benefit. No study gibility for the study, participants were scheduled
has, to date, investigated the impact of expressive for an initial intake with an assessor. All study
writing on sexual problems in any population. procedures were approved by the Institutional
In the current study, we investigated a writing- Review Board of the University of Texas at Austin
based treatment for adult survivors of CSA. Given from 2004 to 2013 and registered on ClinicalTri-
that writing has been shown to affect SSS and that als.gov (identifier NCT01803802).
negative SSS are associated with sexual problems
in women with a history of CSA, a writing inter-
Inclusion and Exclusion Criteria
vention that specifically targets sexual schema may
Women entering the trial had to report at least
be particularly useful in addressing sexual difficul-
one involuntary sexual experience, defined as
ties in this population. To that end, we developed
“unwanted oral, anal, or vaginal intercourse, pen-
a treatment designed to direct participants’ focus
etration of the vagina or anus using objects or
during writing to the impact that sexual abuse may
digits, or genital touching or fondling” before age
have had on their sexual schema, particularly SSS.
16 and no less than 2 years prior to enrollment.
Although written and verbal disclosure appear to
To appropriately measure sexual functioning and
have similar benefits in terms of psychological and
distress, participants were required to either be
physical health outcomes [49], writing uniquely
currently sexually active or be cohabiting in a
offers privacy, which may make it more acceptable
potentially sexual relationship. Additionally, they
for trauma survivors. Moreover, writing as a treat-
had to report sexual dysfunction, distress, or low
ment is simple to administer and cost-effective
sexual satisfaction. The lower age limit was 18;
[50] as it requires minimal input from skilled
there was no upper age limit.
personnel [51].
Women were excluded if they had experienced a
The purpose of the present study was twofold.
traumatic event in the previous 3 months, been a
First, we aimed to test a treatment known to
victim of sexual abuse in the past 2 years, or had
improve depression and PTSD (trauma-focused
been diagnosed with a psychotic disorder in the
writing) for its effects on sexual problems in CSA
previous 6 months. Other psychiatric conditions
survivors. Second, we compared the effects of this
were permissible so long as participant did not
active comparator to a novel focus for expressive
report significant suicidal or homicidal intent at
writing: sexual schema. We had the following
intake. Participants could not be currently receiv-
hypotheses:
ing psychotherapy for sexual or abuse-related con-
1. Women with CSA histories engaging in both cerns; however, participants could be receiving
trauma-focused and sexual schema-focused psychoactive medications if they had been stabi-
expressive writing would exhibit improved lized on those medications for at least 3 months.
levels of depression and PTSD. Participants were excluded if they reported use of
2. Both expressive writing interventions would illicit drugs but were not excluded for alcohol use.
improve sexual dysfunction in women with Women in currently abusive relationships were
CSA histories, but sexual-schema focused also excluded.
expressive writing would improve sexual dys-
function of CSA survivors to a greater extent Sample Characteristics
than would trauma-focused writing. The final sample used in analyses included 91
women with a history of CSA (see Figure 1).
About half of the participants (59%) had been
Method abused by a family member, and the majority
(92%) had at least one penetrative experience. The
Participants majority of participants was white people (64%),
Recruitment married or in a committed relationship (71%),
We recruited participants via newspaper advertise- and had completed at least some college educa-
ments and posts on community websites advertis- tion (78%). Full demographic characteristics are
ing a treatment study for women who had presented in Table 1, and more information on
experienced CSA. Interested women called the lab retention is available at http://bit.ly/wKzXT8.

J Sex Med 2013;10:2177–2189

CSA and Sex Function 2181

Figure 1 Participant flowchart

A detailed analysis of predictors of dropout is completing each assessment session; they were not
available elsewhere [52]. compensated financially for attending treatment
sessions.

Procedure Treatment Sessions

Assessment Sessions There were four study therapists, all women with
Participants completed five 2-hour assessment master’s degrees in psychology. Following their
sessions: pretreatment, posttreatment, 2 weeks pretreatment assessment session, participants were
follow-up, 1 month follow-up, and 6 months scheduled to meet with a study therapist. During
follow-up. Participants met with the same assessor their first treatment session, the therapist offered a
at each assessment session; there were three female brief rationale appropriate to their condition. Each
assessors. At the pretreatment session, participants participant was informed that no one would read
were oriented to study procedures and given infor- her writing until the study was completed. The
mation sufficient to provide informed consent. therapist then administered a brief check-in form
They then completed clinical interviews and ques- that asked the participant about their suicidal ide-
tionnaires on sexual dysfunction, depression, and ation, trauma-related distress, and use of coping
PTSD (see Measures section below). Posttreat- resources since their last visit to the lab. After
ment sessions included the same assessments as addressing any issues from the check-in form,
well as questions regarding perceptions of the therapists oriented the participant to the writing
treatment. Participants were compensated $70 for task for that day, reading the prompt with the

J Sex Med 2013;10:2177–2189

2182 Meston et al.

Table 1 Participant characteristics in final sample to save and close their writing before the therapist
Categorical variables n % returned to the room. Furthermore, participants
were given the option of deleting their writing
Education
Less than high school/GED 3 3.3 before saving and closing it (so that the therapist
Completed high school/GED 13 14.3 would not know if they had saved text or not), or
Some college/college degree 59 64.8 removing their data from analyses after their par-
Advanced degree 12 13.2
Data missing 4 4.4 ticipation; no participant chose this option.
Relationship status Following the writing assignment, the therapist
Single, not dating 12 13.2 briefly evaluated the participant for significantly
Single, dating 10 10.9
In a committed relationship 38 41.8 increased psychological distress related to their
Married 26 28.6 writing or other signs of increased risk of harm to
Data missing 5 5.5 self or others. Safety plans were created as needed,
Ethnicity
Caucasian 58 63.7 including discussions of means of support and
Hispanic/Latina 21 23.1 coping. Women were allowed to leave following
African-American/Black 6 6.6 this risk assessment or to spend the remainder of
Asian-American 4 4.4
Other 7 7.7 the hour talking with the therapist about their
Data missing 4 4.4 writing. A very small minority (approximately 5%)
Current diagnoses of women consistently chose not to stay following
Depression
Yes 15 16.5 the risk assessment. Therapists were not permitted
Subclinical 3 3.3 to conduct any other therapeutic technique (e.g.,
No 66 72.5 cognitive restructuring).
Data missing 1 1.1
PTSD Each additional session followed the same
Yes 19 20.9 format, with a check-in before and after the
Subclinical 17 18.7 30-minute writing period. Treatment was paced
No 47 51.6
Data missing 2 2.2 such that participants were scheduled for no more
Abuser relationship than two sessions per week and never on consecu-
Family 54 59.3 tive days. Most women chose to meet weekly.
Non-family 31 34.1
Data missing 6 6.6 At the end of the fifth treatment session,
Use of psychotropic medications participants were scheduled for a posttreatment
No medication reported 65 71.4 follow-up with the same assessor who completed
Antidepressant(s) only 7 7.7
Antidepressant(s) and other psychoactive 12 13.2 their intake. Following completion of the study, all
medications essays were examined by research assistants for
Other psychoactive medication (e.g., 7 7.7 content relevant to the treatment prompts (includ-
sleep aids)
ing ensuring that women in the trauma condition
Continuous variables M SD wrote about their index sexual trauma); however,
Sexual orientation 2.52 1.854
no essay was judged to be significantly off-prompt
Age 33.7 10.294 [28,53].
Note: The sexual orientation scale is scored such that 0 indicates exclusive
heterosexual attraction and behavior and 6 indicates exclusive homosexual
Conditions
attraction and behavior. This mean indicates most of the sample indicated In keeping with ethical and methodological guide-
predominantly heterosexual attraction but at least some homosexual experi-
ence or interest. Approximately 32% reported a 0, or exclusively heterosexual lines suggested for randomized clinical trials of
attraction and experience, whereas 2% reported a 6, or exclusively homo-
sexual attraction and experience.
psychotherapeutic interventions [54],3 we com-
M = mean; SD = standard deviation pared our experimental treatment (sexual schema-
focused expressive writing) with a known active
treatment (trauma-focused expressive writing).
participant. Participants typed their essays on a
computer into a word document identified by their Sexual Schema-Focused Condition. The schema
unique code and session number.2 condition prompts were developed to focus par-
To ensure privacy, participants were left alone
to write for 30 minutes, and they were instructed 3
Noninferiority designs, in which an experimental treat-
ment is compared against an active comparator, have been
2
If uncomfortable using a computer, participants were recommended by the Food and Drug Administration and
given the option of writing by hand and putting their the American Psychological Association for trials in which
writing into sealed envelopes. Seven participants chose this the safety, cost, and availability of the treatments are neg-
option. ligibly different.

J Sex Med 2013;10:2177–2189

CSA and Sex Function 2183

ticipants’ attention to the impact of their sexual diagnose FSD [57–59]. Participants were consid-
abuse experiences on their thoughts, feelings, and ered “recovered” when they no longer met criteria
beliefs about sexuality (see http://bit.ly/wKzXT8 for this disorder.
for full text of all prompts). The first session
prompt encouraged women to write about how Psychopathology Measures
their sexual abuse may have affected their beliefs Clinician-Administered PTSD Scale (CAPS-
about themselves, sexual partners, or sexuality in 1). Symptoms of PTSD within the last month
general. The second session prompt expanded were assessed with the CAPS-1 [60]. We asked
on the first, asking women to consider the evi- about the most severe trauma experienced (as
dence for and against their beliefs about sex and identified by the Trauma History Questionnaire).
their sexuality. The third and fourth session As recommended in the CAPS manual, we consid-
prompts asked women to consider their reasons ered symptoms as present if the participant scored
for maintaining their sexual beliefs and what would at least 1 for frequency (i.e., once or twice in the
have to change in their lives to change their beliefs. past month) and at least 2 for intensity (i.e., mod-
The final session prompt encouraged women to erate intensity). A total severity score of 45 or
write about their goals for their future sexual life more is considered indicative of clinically relevant
and to focus on their progress and strength. PTSD [61]. Assessors were trained on the CAPS
with the standardized training video developed by
Trauma-Focused Condition. The trauma condition the Department of Veterans Affairs [62] and had
was adapted from the standard expressive writing their assessments reviewed by an experienced psy-
paradigm for this population. The first treatment chometrician. If participants did not report at least
session prompt encouraged participants to write one symptom in each cluster (re-experiencing,
their deepest thoughts and feelings about a trauma hypervigilance, and avoidance), they were classi-
that has affected them, considering how their fied as “subclinical.”
trauma impacted safety, trust, power and control,
and esteem and intimacy (beliefs commonly Structured Clinical Interview for the DSM-IV-
impacted following sexual violence [45,55]). Ses- TR. Current depression and history of major
sions two through four were focused on consider- depressive episodes were assessed with the mood
ing maladaptive beliefs related to the traumatic disorders module of the Structured Clinical Inter-
experience. The final session prompt directed view for the DSM-IV-TR (SCID-1) [63]. Asses-
women to consolidate what they learned during sors were trained according to the training
the previous sessions and to outline goals for the sequence recommended by developers of the
future. SCID (http://bit.ly/XoR9G3) and had their assess-
ments periodically reviewed by the same psycho-
Measures metrician. Participants were classified as
Sexual Functioning Interview. Sexual function was “subclinical” if they met criterion A (depressed
assessed with a structured clinical interview fol- mood or anhedonia) but not criterion B (at least
lowing the criteria for Female Sexual Dysfunction five additional symptoms of depression).
(FSD) presented in the Diagnostic and Statistical
Manual Fourth Edition, Text Revised (DSM- Beck Depression Inventory-II (BDI-II). Symptoms
IV-TR [56]) This interview assessed symptoms of of depression experienced in the past 2 weeks were
hypoactive sexual desire disorder (HSDD), sexual assessed with the BDI-II [64], a widely used and
aversion disorder, female sexual arousal disorder extensively validated 21-item questionnaire.
(FSAD), female orgasm disorder (FOD), vaginis- Scores on the BDI-II can reliably distinguish psy-
mus, and dyspareunia. Each item included an chiatric patients from nondepressed controls as
assessment of presence or absence of each of the well as patients with dysthymia from patients with
symptoms associated with the dysfunction (e.g., major depression [65]. Scores from 0 to 13 indicate
“Do you have a persistent or recurrent lack of no to minimal depression; 14–19 indicate mild
sexual thoughts, fantasies, daydreams, or desire for depression; 20–28 indicate moderate depression;
sexual activity?”), as well as distress related to the and 29–63 indicate severe depression [65].
symptom, time of onset (lifelong or acquired), and
situations in which the symptom was experienced Other Measures
(situational or generalized). This interview has Demographics. A brief demographics questionnaire
been validated and used in previous studies to assessed age, race and ethnicity, sexual orientation

J Sex Med 2013;10:2177–2189

2184 Meston et al.

Table 2 Number of participants meeting criteria for sexual dysfunction across time points
2-week 1-month 6-month Total
Pretreatment Posttreatment follow-up follow-up follow-up recovered (%)
Hypoactive sexual desire disorder
Schema 18 11 9 8 8 56
Trauma 18 15 14 14 14 22
Female sexual arousal disorder
Schema 24 18 11 11 11 54
Trauma 23 20 20 17 17 26
Female orgasmic disorder
Schema 29 22 21 20 20 31
Trauma 27 20 15 15 15 44

and identity, level of education and family income, Results

and type and duration of current romantic Recovery from Sexual Dysfunction
relationship.
We first examined the rate of recovery from clini-
cal diagnoses of sexual dysfunction (HSDD,
Analytic Methods FSAD, and/or FOD; see Table 2). Thirty-six par-
Dichotomous Variables (Diagnosis of ticipants met criteria for HSDD at intake and,
Sexual Dysfunction) over the course of the study, 15 of these partici-
pants experienced recovery (no longer met criteria
To model our dichotomous outcome data, we
for HSDD as determined by a study clinician).
used Cox proportional hazards regressions. The
Condition was a significant predictor of time until
Cox proportional hazards regression tested the
recovery (z = 2.11, P < 0.05; R2 = 0.13). Partici-
incidence of recovery from a sexual dysfunction
pants in the schema condition tended to recover
diagnosis, time until recovery, and whether
more quickly than participants in the trauma con-
incidence and time of recovery differed signifi-
dition (see Figure 2).
cantly between experimental conditions.
Forty-seven participants met criteria for FSAD
at intake, and 21 of these participants recovered
Continuous Variables (Symptoms of Depression at some point in the study. Condition was a
and PTSD) significant predictor of time until recovery
Given the data nonindependence of residuals (z = 2.00, P < 0.05; R2 = 0.09). Again, participants
inherent in longitudinal data, we performed analy- in the schema condition tended to recover more
ses on our continuous outcome variables using quickly than participants in the trauma condition
multilevel Growth Curve Modeling (GCM), avail- (see Figure 3). Fifty-six participants met criteria
able in the hierarchical linear modeling computer for FOD at intake and, of these, 25 women recov-
program (HLMwin v. 6.08 [66]). GCM allowed us ered at some point in the study. Condition was not
to test (i) whether variables changed over time on a significant predictor of time until recovery
average within a sample; (ii) the shape of the (z = -0.56, P value not significant).
average growth rate; (iii) whether there was inter-
individual variation in the shape, strength, or Change in Depression and PTSD
direction of growth rates; and (iv) whether experi- We also tested whether depression and PTSD
mental condition accounted for significant vari- changed on average over time. For depression, we
ance in growth rates. We tested for both quadratic found significant linear and quadratic growth,
and cubic rates of change to capture a number indicating depression improved from pretreat-
of potential patterns, including improvement ment to posttreatment but that there was a slight
from pretreatment to posttreatment with sub- worsening of symptoms over follow-up. Treat-
sequent relapse (single, quadratic curve) and ment condition did not significantly predict post-
improvement from pretreatment to posttreatment treatment scores or rates of change.
with maintenance over follow-up (two-curved, For PTSD symptoms, we found significant
cubic function). For more information about the linear, quadratic, and cubic change (see Figure 4).
construction of our models, see http://bit.ly/ This pattern suggests that, on average, PTSD
wKzXT8. symptoms decreased from pretreatment to

J Sex Med 2013;10:2177–2189

CSA and Sex Function 2185

Figure 2 Proportion of participants

(by condition) meeting criteria for
hypoactive sexual desire disorder
(HSDD) at intake not recovered

posttreatment and that this improvement was minutes was associated with improvements in
maintained over follow-up. Again, treatment con- depression in adult survivors of CSA. There are
dition did not significantly predict scores or rates several possible explanations for this contrast.
of change. First, we spaced out sessions of writing over the
course of several weeks, as opposed to daily ses-
sions as in the protocol of Batten et al. [47]. It is
Discussion
likely that participants need time to process and
Findings from the present study suggest that incorporate changes in their belief systems into
expressive writing was effective in reducing symp- their daily lives, a fact that is reflected in the
toms of depression, PTSD, and sexual dysfunction finding that longer time between writing sessions
in women with a history of CSA. Writing focused is associated with greater improvements [40]. We
on sexual schema was associated with a higher like- also conducted a longer term follow-up (6 months)
lihood of, and faster time to, recovery from HSDD than did Batten et al. (12 weeks); many of the
and FSAD, but not FOD. Each of these findings is improvements observed in the current study
discussed separately below. appeared or were significantly maintained during
In contrast to the findings of Batten et al. [47], the later follow-up assessments. Finally, the par-
we found that five sessions of writing for 30 ticipants in the Batten et al. study were not

Figure 3 Proportion of participants

(by condition) meeting criteria for
female sexual arousal disorder
(FSAD) at intake not recovered

J Sex Med 2013;10:2177–2189

2186 Meston et al.

closure and privacy afforded by the expressive

writing paradigm. Women in the present study
were given absolute choice as to their level of dis-
closure to the study clinicians, which many
reported made it easier to engage more fully in
the treatment. Indeed, one study of expressive
writing in which trauma survivors were required
to read their writing to others in a group setting
showed a worsening of PTSD symptoms [69].
Moreover, writing allowed women to have
control over the pace and intensity of the treat-
ment, a feature that is especially important to
CSA survivors.
It is especially interesting that the sexual schema
condition, which had an ostensibly more special-
ized focus, produced similar decreases in both
depression and PTSD as did the trauma condition.
This suggests that the impact of expressive writing
on maladaptive schema generalized to schema
associated with depression and PTSD, despite the
constraint to focus on sexual schema. In other
words, focusing on sexual schema may have
allowed women to both process sexual issues and
learn skills (e.g., challenging maladaptive cogni-
tions) that then transferred to more general psy-
Figure 4 Change in depression symptoms (BDI) and post-
chopathology. It is also possible that for this
traumatic stress disorder symptoms (CAPS) over time population, depression and PTSD are integrally
linked to sexual schema. Focusing on sexuality,
thus, may have led to changes in psychopathology.
treatment seeking, and thus may have lacked moti- Although it has been suggested that putting con-
vation for change. straints on the topic or directing a focus for expres-
As in several other trials [41], we found that sive writing may distract participants from
expressive writing was associated with reduced expressing themselves freely [34], certain con-
PTSD symptoms. Novel to this study, we saw this straints such as writing about the same trauma
improvement continue well into a long-term during each writing session rather than different
follow-up with a CSA population. Several studies traumas each time have been shown to improve
that documented no effect of expressive writing on long-term gains [70]. Certainly in our study,
PTSD used undergraduate student samples, not writing about sexual schema did not reduce the
clinical populations [67]. PTSD has been termed a general mental health benefits that were also seen
“disorder of recovery” because the difference in the traditional trauma-focused writing but also
between normative and pathological responses to conferred additional benefits in improving sexual
trauma lay not in the symptoms (e.g., hypervigi- function. It would be helpful to compare specific
lance) but in the lack of resolution of those symp- prompts to free writing in future studies.
toms [68]. By definition, PTSD is resistant to Although both writing groups improved in
spontaneous improvement. On average, partici- sexual function, greater gains were seen in the
pants in our sample presented with subclinical sexual schema condition. Because their sexual
PTSD. However, for the most part, these women trauma happened so early in life, survivors of CSA
had been experiencing significant levels of symp- must not only resolve the traumatic memory itself,
toms since childhood, and it is notable that such but the contingent effect on the rest of their sexual
resistant symptoms were reduced to the nonclini- development. It is likely that women in the sexual
cal range and remained in remission at a 6-month schema condition were better able to focus on the
follow-up. whole of their sexual identity and belief system,
It is possible that the key to reducing resistant including how the trauma impacted them, than
symptoms lay in the unique combination of dis- those in the trauma condition who considered only

J Sex Med 2013;10:2177–2189

CSA and Sex Function 2187

how the trauma impacted them. Also, our findings Expressive writing is a treatment that is very
suggest that changes in sexual well-being were easy to conduct, easily accessible, generally accept-
not entirely dependent on changes in depression able and intuitive to most patients, and cost-
and PTSD, as both treatment groups showed effective as it requires minimal trained personnel
similar changes in psychopathology but differen- to administer. The present study suggests that it
tial changes in sexual function. may also confer unique benefits to women with a
To our knowledge, only one other psychological history of CSA: namely, it appears to improve
treatment has been shown to improve sexual func- sexual function in this population.
tion in women with a history of CSA [20]. In the
case of Brotto et al., post hoc follow-ups revealed
that women with a history of CSA benefited more Acknowledgments
from a mindfulness-based educational group than This research was supported by grant number 1 RO1
women without such histories. Given this trend, it HD051676 from the National Institute of Child Health
is possible that expressive writing may also have a and Human Development (NICHD) to Cindy M.
lesser effect on sexual dysfunction in women Meston. Its contents are solely the responsibility of the
without a history of CSA than that seen in the authors and do not necessarily represent the official
present study. However, as effects of expressive views of the NICHD.
writing are stronger in physical than psychological Corresponding Author: Cindy M. Meston, PhD,
health outcomes [39], it is possible that effects on Department of Psychology, University of Texas at
sexual function in a population without psycho- Austin, 108 E. Dean Keeton, Austin, TX 78712, USA.
logical distress would be comparable. Tel: (512) 232-4644; Fax: (512) 471-5395; E-mail:
The present study had limitations that, if [email protected]
addressed in future studies, may lead to an even
Conflict of Interest: The authors report no conflicts of
greater understanding of the generalizability and interest.
mechanisms of action of the expressive writing
treatments. We did not study men with CSA his-
tories; however, as meta-analyses on expressive Statement of Authorship
writing paradigms suggest that the beneficial
Category 1
effects on health are greater for men than for
(a) Conception and Design
women [40], we would expect the effects of the
Cindy M. Meston
treatments to be even larger for male CSA survi- (b) Acquisition of Data
vors. For ethical reasons, the recruitment materials Cindy M. Meston; Tierney A. Lorenz
for the present study explicitly stated this was a (c) Analysis and Interpretation of Data
treatment study for survivors of CSA. There is Cindy M. Meston; Tierney A. Lorenz; Kyle R.
some evidence that the nature of sexual problems Stephenson
differ for women who have experienced abusive
situations but do not identify themselves as having Category 2
been abused vs. those who identify as survivors [4]. (a) Drafting the Article
As such, we do not know if the present findings Cindy M. Meston; Tierney A. Lorenz; Kyle R.
would extend to a population that does not identify Stephenson
as abused. Similarly, the findings presented here (b) Revising It for Intellectual Content
may not apply to a different population than that Cindy M. Meston; Tierney A. Lorenz; Kyle R.
studied, which was on average white, college- Stephenson
educated, and partnered. The present study used a
therapist-guided approach, in which patients had Category 3
the option of exploring what they wrote with an (a) Final Approval of the Completed Article
empathic but objective observer; it would be useful Cindy M. Meston; Tierney A. Lorenz; Kyle R.
to see if writing with no therapist contact would Stephenson
confer the same results. If so, it would be simple to
integrate expressive writing instructions into edu- References
cation provided to CSA survivors by primary care
1 Leonard LM, Follette VM. Sexual functioning in women
facilities—that is, the most common settings in reporting a history of child sexual abuse: Review of the empiri-
which women’s sexual health concerns are first cal literature and clinical implications. Annu Rev Sex Res
addressed [71]. 2002;13:346–88.

J Sex Med 2013;10:2177–2189

2188 Meston et al.

2 Kinzl JF, Traweger C, Biebl W. Sexual dysfunctions: Relation- 23 Rellini AH. Review of the empirical evidence for a theoretical
ship to childhood sexual abuse and early family experiences in model to understand the sexual problems of women with a
a nonclinical sample. Child Abuse Negl 1995;19:785–92. history of CSA. J Sex Med 2008;5:31–46.
3 Sarwer DB, Durlak JA. Childhood sexual abuse as a predictor 24 Impett EA, Tolman DL. Late adolescent girls sexual experi-
of adult female sexual dysfunction: A study of couples seeking ences and sexual satisfaction. J Adolesc Res 2006;21:628–46.
sex therapy. Child Abuse Negl 1996;20:963–72. 25 Gates E, Galask R. Psychological and sexual functioning in
4 Rellini AH, Meston CM. Sexual function and satisfaction in women with vulvar vestibulitis. J Psychosom Obstet Gynaecol
adults based on the definition of child sexual abuse. J Sex Med 2001;22:221–8.
2007;4:1312–21. 26 Cyranowski JM, Aarestad SL, Andersen BL. The role of sexual
5 Nelson EC, Heath AC, Madden PAF, Cooper ML, Dinwiddie self-schema in a diathesis-stress model of sexual dysfunction.
SH, Bucholz KK, Glowinski A, McLaughlin T, Dunne MP, Appl Prev Psychol 1999;8:217–28.
Statham DJ. Association between self-reported childhood 27 Meston CM, Rellini AH, Heiman JR. Women’s history of
sexual abuse and adverse psychosocial outcomes. Arch Gen sexual abuse, their sexuality, and sexual self-schemas. J Consult
Psychiatry 2002;59:139–45. Clin Psychol 2006;74:229–36.
6 U.S. Department of Health and Human Services. Child Mal- 28 Lorenz T, Meston CM. Associations among childhood sexual
treatment 2009. Washington, DC: U.S. Department of Health abuse, language use, and adult sexual functioning and satisfac-
and Human Services, Administration for Children and Fami- tion. Child Abuse Negl 2012;36:190–9.
lies, Administration on Children, Youth and Families, Chil- 29 Rellini AH, Meston CM. Sexual desire and linguistic analysis:
dren’s Bureau. 2010:vii–x. A comparison of sexually-abused and non-abused women.
7 Beck JJ, Bekker MD, van Driel MF, Roshani H, Putter H, Arch Sex Behav 2007;36:67–77.
Pelger R, Elzevier HW. Prevalence of sexual abuse among 30 Rellini AH, David A, Meston CM. Implicit and explicit cog-
patients seeking general urological care. J Sex Med nitive sexual processes in survivors of childhood sexual abuse. J
2011;8:2733–8. Sex Med 2011;8:3098–107.
8 Maltz W. Sexual healing from sexual abuse. SIECUS Rep 31 Niehaus AF, Jackson J, Davies S. Sexual self-schemas of female
2000;29:17–23. child sexual abuse survivors: Relationships with risky sexual
9 Berman LA, Berman JR, Bruck D, Pawar RV, Goldstein I. behavior and sexual assault in adolescence. Arch Sex Behav
Pharmacotherapy or psychotherapy?: Effective treatment for 2010;39:1359–74.
FSD related to unresolved childhood sexual abuse. J Sex 32 Reissing E, Binik Y, Khalifè S, Cohen D, Amsel R. Etiological
Marital Ther 2001;27:421–5. correlates of vaginismus: Sexual and physical abuse, sexual
10 Maltz W. Treating the sexual intimacy concerns of sexual knowledge, sexual self-schema, and relationship adjustment.
abuse survivors. Sex Relation Ther 2002;17:321–7. J Sex Marital Ther 2003;29:47–59.
11 Harrop-Griffith J, Katon W, Walker E, Holm L, Russo J, 33 Lemieux SR, Byers ES. The sexual well-being of women who
Hickok L. The association between chronic pelvic pain, psy- have experienced child sexual abuse. Psychol Women Q
chiatric diagnoses, and childhood sexual abuse. Obstet 2008;32:126–44.
Gynecol 1988;71:589–94. 34 Pennebaker JW, Chung CK. Expressive writing: Connections
12 Noll JG, Trickett PK, Putnam FW. A prospective investiga- to physical and mental health. In: Friedman HS, ed. Oxford
tion of the impact of childhood sexual abuse on the develop- handbook of health psychology. New York: Oxford University
ment of sexuality. J Consult Clin Psychol 2003;71:575–86. Press; 2011:417–37.
13 Maniglio R. The impact of child sexual abuse on health: A 35 Park CL, Blumberg CJ. Disclosing trauma through writing:
systematic review of reviews. Clin Psychol Rev 2009;29:647– Testing the meaning-making hypothesis. Cognit Ther Res
57. 2002;26:597–616.
14 Davis JL, Petretic-Jackson PA, Ting L. Intimacy dysfunction 36 O’Connor DB, Hurling R, Hendrickx H, Osborne G, Hall J,
and trauma symptomatology: Long-term correlates of differ- Walklet E, Whaley A, Wood H. Effects of written emotional
ent types of child abuse. J Trauma Stress 2001;14:63–79. disclosure on implicit self esteem and body image. Br J Health
15 Rieckert J, Möller AT. Rational-emotive behavior therapy Psychol 2011;16:488–501.
in the treatment of adult victims of childhood sexual abuse. 37 Cameron LD, Nicholls G. Expression of stressful experiences
J Ration Emot Cogn Behav Ther 2000;18:87–101. through writing: Effects of a self-regulation manipulation for
16 Follette VM. Marital therapy for sexual abuse survivors. New pessimists and optimists. Health Psychol 1998;17:84–92.
Dir Ment Health Serv 1991;1991:61–71. 38 Lewis RJ, Derlega VJ, Clarke EG, Kuang JC, Jacobs AM,
17 Classen CC, Palesh OG, Cavanaugh CE, Koopman C, Kaupp McElligott MD. An expressive writing intervention to cope
JW, Kraemer HC, Aggarwal R, Spiegel D. A comparison of with lesbian-related stress: The moderating effects of openness
trauma-focused and present-focused group therapy for survi- about sexual orientation. Psychol Women Q 2005;29:149–57.
vors of childhood sexual abuse: A randomized controlled trial. 39 Frattaroli J. Experimental disclosure and its moderators: A
Psychol Trauma 2011;3:84–93. meta-analysis. Psychol Bull 2006;132:823–65.
18 Hazzard A, Rogers JH, Angert L. Factors affecting group 40 Smyth JM. Written emotional expression: Effect sizes,
therapy outcome for adult sexual abuse survivors. Int J Group outcome types, and moderating variables. J Consult Clin
Psychother 1993;43:453–68. Psychol 1998;66:174–84.
19 Hébert M, Bergeron M. Efficacy of a group intervention for 41 Sloan DM, Marx BP. A closer examination of the structured
adult women survivors of sexual abuse. J Child Sex Abus written disclosure procedure. J Consult Clin Psychol
2007;16:37–61. 2004;72:165–75.
20 Brotto L, Basson R, Luria M. A mindfulness based group 42 Ullrich PM, Lutgendorf SK. Journaling about stressful events:
psychoeducational intervention targeting sexual arousal disor- Effects of cognitive processing and emotional expression. Ann
der in women. J Sex Med 2008;5:1646–59. Behav Med 2002;24:244–50.
21 Andersen BL, Cyranowski JM. Women’s sexual self-schema. 43 Klein K, Boals A. Expressive writing can increase working
J Pers Soc Psychol 1994;67:1079–100. memory capacity. J Exp Psychol Gen 2001;130:520–33.
22 Catlin G, Epstein S. Unforgettable experiences: The relation 44 Graybeal A, Sexton JD, Pennebaker JW. The role of story-
of life events to basic beliefs about self and world. Soc Cogn making in disclosure writing: The psychometrics of narrative.
1992;10:189–209. Psychol Health 2002;17:571–81.

J Sex Med 2013;10:2177–2189

CSA and Sex Function 2189

45 Resick PA, Galovski TE, Uhlmansiek MOB, Scher CD, Clum 58 Meston CM, Rellini AH, Telch MJ. Short-and long-term
GA, Young-Xu Y. A randomized clinical trial to dismantle effects of Ginkgo biloba extract on sexual dysfunction in women.
components of cognitive processing therapy for posttraumatic Arch Sex Behav 2008;37:530–47.
stress disorder in female victims of interpersonal violence. J 59 Seal BN, Meston CM. The impact of body awareness on sexual
Consult Clin Psychol 2008;76:243–58. arousal in women with sexual dysfunction. J Sex Med 2007;4:
46 Kearns MC, Edwards KM, Calhoun KS, Gidycz CA. Disclo- 990–1000.
sure of sexual victimization: The effects of Pennebaker’s emo- 60 Blake DD, Weathers FW, Nagy LM, Kaloupek DG,
tional disclosure paradigm on physical and psychological Klauminzer G, Charney DS, Keane TM. A clinician rating
distress. J Trauma Dissociation 2010;11:193–209. scale for assessing current and lifetime PTSD: The CAPS-1.
47 Batten SV, Follette VM, Rasmussen Hall ML, Palm KM. Behav Ther 1990;13:187–8.
Physical and psychological effects of written disclosure among 61 Weathers FW, Ruscio AM, Keane TM. Psychometric proper-
sexual abuse survivors. Behav Ther 2003;33:107–22. ties of nine scoring rules for the Clinician-Administered Post-
48 Koopman C, Ismailji T, Holmes D, Classen CC, Palesh O, traumatic Stress Disorder Scale. Psychol Assess 1999;11:
Wales T. The effects of expressive writing on pain, depression 124–33.
and posttraumatic stress disorder symptoms in survivors of 62 Department of Veterans Affairs. Clinician Administered
intimate partner violence. J Health Psychol 2005;10:211–21. PTSD Scale (CAPS) for DSM-IV. In: Development
49 Esterling BA, Antoni MH, Fletcher MA, Margulies S, Schnei- DoVAHSRa, ed. Washington, DC: Department of Veterans
derman N. Emotional disclosure through writing or speaking Affairs; 2001.
modulates latent Epstein-Barr virus antibody titers. J Consult 63 First M, Spitzer R, Gibbon M, Williams J. Structured clinical
Clin Psychol 1994;62:130–40. interview for DSM-IV-TR axis I disorders, research version,
50 Pennebaker JW. Theories, therapies, and taxpayers: On the patient edition (SCID-I/P). New York: Biometrics Research,
complexities of the expressive writing paradigm. Clin Psychol New York State Psychiatric Institute; 2002.
2004;11:138–42. 64 Beck AT, Steer RA, Brown GK. Manual for the BDI-II. San
51 Lange A, Rietdijk D, Hudcovicova M, Van De Ven JP, Antonio, TX: Psychological Corporation; 1996.
Schrieken B, Emmelkamp PMG. Interapy: A controlled ran- 65 Beck AT, Steer RA, Carbin MG. Psychometric properties
domized trial of the standardized treatment of posttraumatic of the Beck Depression Inventory: Twenty-five years of
stress through the internet. J Consult Clin Psychol 2003;71: evaluation. Clin Psychol Rev 1988;8:77–100.
901–9. 66 Raudenbush SW, Bryk AS, Cheong YF, Congdon RT. HLM
52 Harte CB, Hamilton LD, Meston CM. The effects of an 6: Hierarchical linear and nonlinear modeling. 6.08 edition.
expressive writing intervention for adult women survivors of Chicago, IL: Scientific Software International; 2004.
childhood sexual abuse: Comparisons between completers and 67 Sloan DM, Marx BP, Greenberg EM. A test of written emo-
dropouts. J Child Sex Abus in press. tional disclosure as an intervention for posttraumatic stress
53 Lorenz T, Pulverman C, Meston CM. Sudden gains during disorder. Behav Res Ther 2011;49:299–304.
patient-directed expressive writing treatment predicts depres- 68 Shalev A. PTSD: A disorder of recovery. In: Kirmayer L,
sion reduction in women with history of childhood sexual Lemelson R, Barad M, eds. Understanding trauma: Biological,
abuse. Cognit Ther Res; doi: 10.1007/s10608-012-9510-3. psychological and cultural perspectives. Cambridge, UK:
54 La Vaque T, Rossiter T. The ethical use of placebo controls in Cambridge University Press; 2007:207–23.
clinical research: The Declaration of Helsinki. Appl Psycho- 69 Gidron Y, Peri T, Connolly JF, Shalev AY. Written emotional
physiol Biofeedback 2001;26:23–37. disclosure in posttraumatic stress disorder: Is it beneficial for
55 Resick PA, Schnicke MK. Cognitive processing therapy for the patient? J Nerv Ment Dis 1996;184:505–7.
rape victims: A treatment manual. Newbury Park, CA: Sage 70 Sloan DM, Marx BP, Epstein EM. Further examination
Publications; 1993. of the exposure model underlying the efficacy of written
56 American Psychiatric Association. Diagnostic and statistical emotional disclosure. J Consult Clin Psychol 2005;73:
manual of mental disorders. Revised 4th edition. Washington, 549–54.
DC: Author; 2000. 71 Mercer CH, Fenton KA, Johnson AM, Wellings K, Macdowall
57 Meston CM, Rellini AH, McCall K. The sensitivity of con- W, McManus S, Nanchahal K, Erens B. Sexual function prob-
tinuous laboratory measures of physiological and subjective lems and help seeking behaviour in Britain: National probabil-
sexual arousal for diagnosing women with sexual arousal ity sample survey. BMJ 2003;327:426.
disorder. J Sex Med 2010;7:938–50.

J Sex Med 2013;10:2177–2189

Copyright of Journal of Sexual Medicine is the property of Wiley-Blackwell and its content
may not be copied or emailed to multiple sites or posted to a listserv without the copyright
holder's express written permission. However, users may print, download, or email articles for
individual use.