Crosby Jesse M Acceptance and Commitment Therapy For 2016
Crosby Jesse M Acceptance and Commitment Therapy For 2016
Crosby Jesse M Acceptance and Commitment Therapy For 2016
Acceptance and Commitment Therapy for Problematic Internet Pornography
Use: A Randomized Trial
PII: S0005-7894(16)00017-4
DOI: doi: 10.1016/j.beth.2016.02.001
Reference: BETH 609
Please cite this article as: Crosby, J.M. & Twohig, M.P., Acceptance and Commitment
Therapy for Problematic Internet Pornography Use: A Randomized Trial, Behavior Ther-
apy (2016), doi: 10.1016/j.beth.2016.02.001
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 1
A Randomized Trial
PT
Jesse M. Crosby
RI
McLean Hospital / Harvard Medical School
&
SC
Utah State University
NU
Michael P. Twohig
MA
Utah State University
D
P TE
Corresponding Author:
CE
Email: [email protected]
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 2
Abstract
Problematic internet pornography use is the inability to control the use of pornography,
the experience of negative cognitions or emotions regarding pornography use, and the resulting
PT
negative effects on quality of life or general functioning. This study compared a 12-session
RI
individual protocol of acceptance and commitment therapy (ACT) for problematic internet
pornography use to a waitlist control condition with 28 adult males, all but one of whom were
SC
members of the Church of Jesus Christ of Latter-day Saints. Measures of self-reported
NU
pornography viewing, standardized measures of compulsive sexual behavior and related
cognitions, and quality of life occurred at pretreatment, posttreatment, and three-month follow-
MA
up. Results demonstrate significant between-condition reductions in pornography viewing
compared to the waitlist condition (93% reduction ACT vs. 21% waitlist). When combining all
D
TE
participants (N=26), a 92% reduction was seen at posttreatment and an 86% reduction at three-
month follow-up. Complete cessation was seen in 54% of participants at posttreatment and at
P
least a 70% reduction was seen in 93% of participants. At the three-month follow-up assessment,
CE
35% of participants showed complete cessation, with 74% of participants showing at least 70%
AC
A Randomized Trial
PT
problematic or compulsive sexual behavior, and much of the focus of this research has been on
RI
how best to conceptualize and diagnose the behavior. Conceptualizations have included sexual or
pornography addiction (Hilton Jr & Watts, 2011; Orzack & Ross, 2000), sexual impulsivity
SC
(Mick & Hollander, 2006), compulsive sexual behavior (Coleman, 1991), sexual compulsivity
NU
(Cooper, Putnam, Planchon, & Boies, 1999), out-of-control sexual behavior (Salisbury, 2008),
and hypersexual behavior or hypersexuality (Rinehart & McCabe, 1998) which was proposed,
MA
but ultimately not included in, the Diagnostic and Statistical Manual of Mental Disorders-5
(Reid et al., 2012). Three formal classes of disorders have been used to provide criteria and
D
TE
(Schneider, 1994), (b) impulse control disorders (Grant & Potenza, 2010), and (c) obsessive-
P
Problematic internet pornography is generally defined by the inability to control the use
AC
and the resulting negative effects on quality of life or general functioning (Coleman, Miner,
Ohlerking, & Raymond, 2001; McBride, Reece, & Sanders, 2008; Reid, 2007). This could
include damaged relationships, loss of productivity, impaired performance at work or school, job
loss, financial expenses, guilt/shame, personal distress, and other forms of psychopathology
(McBride et al., 2008). Additionally, problematic pornography use has been identified as a major
contributing factor to marital separation and divorce (Dedmon, 2002; Schneider, 2000). Similar
to other clinical behaviors, the use of pornography is not viewed as inherently problematic. It is
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 4
problematic only to the extent to which it becomes excessive and leads to problematic emotional,
PT
compulsive sexual behavior (Kuzma & Black, 2008), and that approximately half of these
RI
individuals have an interest in pornography as part of their compulsive behavior (Black,
Kehrberg, Flumerfelt, & Schlosser, 1997), leading to an estimate of about 1.5% to 3% of the US
SC
general adult population. With the increased use of the internet, that number is likely an
NU
underestimate. For example, a study of over 9000 internet users found that between 9% and 15%
of the participants reported distress related to their use of the internet for sexual purposes and
MA
10% reported their behavior as “addictive” (Cooper, Delmonico, Griffin-Shelley, & Mathy,
2004). In a survey of over 9000 individuals who had accessed pornographic or sexual content on
D
TE
the internet, 17% scored in the problematic range for sexual compulsivity (Cooper, Delmonico,
& Burg, 2000). In another survey of males involved in online sexual activity, 6.5% reported
P
Delmonico, & Mathy, 2001). In a study of Swedish men and women, 5% of women and 13% of
AC
men reported some problems with sexual internet use and 2% of women and 5% of men reported
serious problems with sexual internet use (Ross, Månsson, & Daneback, 2012). There is
particular concern about the prevalence of these behaviors among adolescents, where general
internet use is high and it is assumed that problematic sexual use of the internet occurs for a
percentage of that general sample (Owens, Behun, Manning, & Reid, 2012).
treatments that have been suggested or investigated include motivational interviewing (Del
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 5
Giudice & Kutinsky, 2007), cognitive behavior therapy (Young, 2007), 12-step programs
(Schneider, 1994), and emotion-focused therapy (Reid & Woolley, 2006). Although these
recommendations are promising, they are not supported with controlled outcome work and the
PT
uncontrolled work that does exist is often with variants of problematic internet pornography use
RI
(e.g., compulsive internet use, relationship problems from viewing; Winkler et al., 2013).
Current research on the way inner experiences (i.e., thoughts, emotions, physical
SC
sensations) are addressed and function may have important implications for the understanding
NU
and treatment of problematic pornography use. Multiple studies have shown that the way one
interacts with urges to view pornography affects not only the rate of viewing, but emotional
MA
distress from viewing and problems related to viewing (Levin, Lillis, & Hayes, 2012; Twohig et
al., 2009). Struggling with thoughts is an important part of how sexual compulsivity is defined
D
TE
and measured (Reid, Bramen, Anderson, & Cohen, 2013). Indeed, a commonly used measure of
sexual compulsivity (Cognitive and Behavioral Outcomes of Sexual Behavior Scale; McBride et
P
al., 2008) assesses individuals' difficulty controlling sexual thoughts and behaviors (Kalichman
CE
Given that attempts to regulate certain thoughts and urges are not helpful in the long run,
research has focused on the utility of acceptance and mindfulness-based procedures (Levin,
Hildebrandt, Lillis, & Hayes, 2012), especially when applied to intrusive thoughts (Marcks &
instead of attempts to regulate them. This is the focus of acceptance and commitment therapy
(ACT; Hayes, Strosahl, & Wilson, 1999), and suggests that ACT may be an effective treatment
for problematic pornography use. ACT targets processes that generally aim to decrease the
effects of inner experiences (e.g., urges to use pornography) on overt behavior and increase the
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 6
effects of other inner experiences (e.g., studies investigating values-based work or defusion in
isolation) on behavior. This is consistent with a recent study showing that lack of mindfulness is
positively related to hypersexuality beyond emotion dysregulation, impulsivity, and stress (Reid
PT
et al., 2013).
RI
There is a growing body of research in support of ACT for the treatment of a wide variety
of disorders (Hayes, Luoma, Bond, Masuda, & Lillis, 2006), including disorders to which
SC
problematic pornography use is most often compared (OCD, impulse control disorders, and
NU
substance use disorders). The specific ACT processes of change are also supported in component
studies (e.g., studies of values or defusion only), showing that improvement of those processes
MA
outside of a large treatment package is clinically useful (Levin, Hildebrandt, et al., 2012). The
case for applying ACT to problematic internet pornography use is strengthened further by
D
TE
preliminary work with ACT for problematic pornography use in a single subjects design study
(Twohig & Crosby, 2010). This study tested the effectiveness of ACT with six adult male
P
participants treated in eight 1.5-hour sessions. Treatment resulted in an 85% reduction in viewing
CE
at posttreatment with results being maintained at 3-month follow-up (83% reduction). These
AC
findings, coupled with the lack of treatment outcome work in this area, suggest the need for a
Method
Participants
community, and announcements in university classes. Participants were eligible to enroll if they
met criteria for problematic pornography use established from preliminary investigations as no
formal diagnostic criteria have been established (Twohig & Crosby, 2010). The criteria for
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 7
inclusion were: (a) the individual must have engaged in problematic internet pornography use for
more than 6 months; (b) the individual must have viewed pornography with a frequency of at
least two sessions per week, on average, for the month previous to enrolling in the study; (c) the
PT
individual must have experienced significant distress and/or functional impairment in his life;
RI
and (d) the individual must have had at least one unsuccessful attempt at stopping the behavior.
Participants were ineligible if they (a) were currently receiving psychotherapy; (b) started,
SC
changed, or were planning to change a psychotropic medication within 30 days; (c) were not
NU
capable of participating in the research due to physical/medical complications; (d) met criteria
for substance dependence; or (e) had been diagnosed with an intellectual or developmental
MA
disability. Eligibility was assessed using a semi-structured interview to assess all inclusion and
exclusion criteria for the study. In addition, an abbreviated version of the Structured Clinical
D
TE
Interview for DSM-IV Axis I Disorders, Clinician Version (SCID-C; First, Spitzer, Gibbon, &
Twenty-eight participants met eligibility requirements for participation in the study. All
participants were male with an average age of 29.3 (SD=11.4). Most participants were married
AC
(54%) and 92% were white American. Even though participants were recruited from the
community of a medium sized city, all but one participant were members of the Church of Jesus
Christ of Latter-day Saints. This is likely the result of the majority of citizens in the town being
members of that church, and pornography being against their religious beliefs. On average, they
had struggled with pornography use for 13.6 years (SD=11.8). All participants reported viewing
pornography on the internet, and others additionally reported viewing it in magazines (n=3), via
cable or satellite television (n=5), or video rental (n=1). Other problematic sexual actions were
as follows: phone/internet sex with strangers (n=3), strip clubs (n=1), infidelity (n=1), and hiring
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 8
of prostitutes (n=1). All had tried to stop viewing with many attempting individual
psychotherapy (n=13), followed by group therapy (n=4), 12-step programs (n=7), religious
counseling (n=6), and using self-help (n=2). According to the SCID-C, comorbid diagnoses were
PT
as follows: major depressive disorder (n=5), bipolar disorder (n=1), generalized anxiety disorder
RI
(n=2), and OCD (n=1). Use of psychotropic medications were as follows: antidepressants (n=3),
mood stabilizers (n=1), and antipsychotics (n=1). Participants were randomly assigned to one of
SC
two conditions (n=14 ACT, n=14 wait list). Results showed no statistical difference between the
NU
ACT and the control conditions on any pretreatment variables. Participant flow is reported in
Figure 1.
MA
Measures
age, religion, ethnicity/race, years of pornography viewing, how they accessed pornography,
other problematic sexual behaviors, previous treatment attempts for problematic pornography
P
Pornography Viewing Questionnaire (DPVQ), which was developed for purposes of this study.
The DPVQ is a modified version of the Daily Drinking Questionnaire (DDQ; Collins, Parks, &
Marlatt, 1985), an instrument originally developed to measure the quantity of alcohol use as well
as hours spent engaging in alcohol consumption. The format of the DDQ improves the accuracy
of self report and has been shown to be valid and reliable (Baer, Stacy, & Larimer, 1991;
Neighbors, Lee, Lewis, Fossos, & Larimer, 2007). Our modified version, the DPVQ, was similar
in format to the DDQ but was instead used to measure weekly sessions of pornography viewing
and the number of hours spent viewing pornography for each day of the week, as well as
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 9
other items not reported in this manuscript. The daily results were reported to the therapist at
each weekly session. Weekly hour totals were obtained by summing the results from each
weekday.
PT
Sexual Compulsivity Scale (SCS). The SCS (Kalichman et al., 1994) consists of 10
RI
items designed to assess sexual compulsivity. Seven of the items address sexual desires and how
an individual reacts to them, and three of the items focus more on the negative effects of sexual
SC
thoughts and behaviors. The SCS has been shown to have adequate validity and reliability
NU
(Dodge, Reece, Cole, & Sandfort, 2004; Kalichman et al., 1994; Kalichman & Rompa, 1995;
Perry, Accordino, & Hewes, 2007). It is internally consistent (α = .86) and has test-retest
MA
reliability of .64. The SCS has also been found to be predictive of internal difficulties (e.g.,
loneliness, low self-esteem, and beliefs about self-control; Kalichman et al., 1994) and lack of
D
TE
intention to change potentially problematic sexual behaviors (Kalichman & Rompa, 1995). The
SCS has been shown to predict problematic sexual behaviors, and it predicts internal events that
P
are conducive to engaging in sexual behaviors that are likely to be problematic. Cronbach’s
CE
Subscale. The CBOSB (McBride et al., 2008) consists of 20 items measuring worries about
negative consequences of sexual practices over the last year. The CBOSB has demonstrated
adequate reliability and validity (McBride et al., 2008). Internal consistency for the CBOSB
cognitive scale has been high (α = .89). The construct validity of the subscale has been
demonstrated using a principal component analysis in which the six factors/subscales explained
Quality of Life Scale (QOLS). The QOLS (Burckhardt & Anderson, 2003) is a 16-item
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 10
scale that measures how satisfied people are with the quality of their lives in several areas (e.g.,
relationships, employment, health, recreation), and has been used in a variety of populations. The
items are rated on a 1-7 point scale where 1 = “terrible,” and 7 = “delighted.” Convergent and
PT
discriminant validity was shown in that the QOLS was highly correlated with a measure of life
RI
satisfaction (r=.68 to.75), but had lower correlations with a measure of physical health (r=.25 to
.48). The measure is scored by summing the scores for all 16 items. It has been found to be
SC
internally consistent (α = .89 to .92), and has demonstrated temporal stability (r = .78 to .84) over
NU
three weeks. Cronbach’s alpha at pretreatment was .86.
Procedures
MA
The effects of the treatment were assessed through a randomized controlled trial with a
waitlist control condition, with the waitlist condition receiving treatment after the waiting period.
D
TE
Follow-up assessments were collected three months after treatment. Participants responded to
the recruitment efforts by phone or email and any questions or concerns regarding the study were
P
answered. After agreeing to participate in the study, participants attended a 2-hour pretreatment
CE
session during which they were provided with an Institutional Review Board -approved informed
AC
consent for their review and signature. This initial session also included the interviews to assess
for eligibility and the administration of the measures to gather information on the participants’
backgrounds, relevant difficulties, and other information pertaining to the research questions. To
assess the effectiveness of ACT, the assessments were administered at posttreatment for the ACT
condition and after a 12-week waiting period for the waitlist condition. The waitlist condition
was offered treatment after their second assessment, and posttreatment assessment was
completed after treatment was completed. All participants completed a follow-up assessment 12
either an ACT or a waitlist condition. Analyses showed no statistical difference between ACT
and the control conditions on age, education level, marital status, years of viewing, hours
PT
viewing per week, or any of the dependent measures. Participants in the ACT condition began
RI
treatment immediately after randomization. Participants in the waitlist condition began treatment
after 12 weeks had elapsed and they had completed the post-waitlist assessment.
SC
Treatment
NU
ACT for problematic pornography use. A 10-session ACT manual for problematic
pornography use (Twohig & Crosby, 2010) was modified and expanded for this study.
MA
Treatment consisted of 12 individual weekly 1-hour sessions of ACT. The goals of this treatment
protocol were: (a) to help the client determine effective strategies for responding to urges to
D
TE
engage in the pornography use, (b) to practice using these strategies outside of session, (c) to
gradually decrease pornography use, and (d) to increase occurrence of high quality of life
P
activities. Table 1 provides a summary of the treatment components and specific interventions
CE
used. A copy of the complete manual is available from the corresponding author.
AC
Treatment Adherence. The intervention was provided by the first author (23
participants). Both graduate student therapists were supervised by the second author, who is a
licensed psychologist. All of the treatment sessions were recorded (video and audio) to monitor
treatment integrity. A sample of the sessions (68 of 315, 21.59%) was selected to be viewed and
scored for treatment integrity using a standardized treatment integrity scoring system used in
previous ACT research (Plumb & Vilardaga, 2010). The sessions to be reviewed were selected
systematically and objectively so that of the 12 total sessions, approximately three sessions from
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 12
each participant and six of each session number were reviewed. The review of the sessions was
conducted by three trained graduate students who showed high reliability (.90 or greater) with
the senior (MPT) assessor on two consecutive videos. The graduate students all worked in a
PT
laboratory that focuses on ACT, but in addition, three two-hour sessions on how to score
RI
occurred prior to testing on the assessing protocol.
In the standardized scoring system, sessions were scored for both the quantity and the
SC
quality of coverage of each ACT process on a 5-point Likert scale. For quantity of coverage, 1 =
NU
the process was never explicitly covered, 2 = the process occurred at least once and not in an in-
depth manner, 3 = the process occurred several times and was covered at least once in a
MA
moderately in-depth manner, 4 = the process occurred with relatively high frequency and was
addressed in a moderately in-depth manner, and 5 = the process occurred with high frequency
D
TE
and was covered in a very in-depth manner. The review showed that all of the ACT processes
were rated as a 5 in at least one of the reviewed sessions, indicating that all of the ACT processes
P
were covered in depth in at least one session. The mean ratings for each process over all 12
CE
sessions were: acceptance = 3.50 (SD = 1.24), defusion = 3.76 (SD = 1.40), self as context = 1.68
AC
(SD = 1.20), contact with the present moment = 1.62 (SD = 1.04), values = 2.37 (SD = 1.53), and
committed action = 2.40 (SD = 1.43). Sessions were also reviewed for intervention techniques
that are inconsistent with the ACT model, including challenging cognitions, suggesting that
thoughts or feelings can cause behavior, and behavioral management strategies to avoid triggers
of private events. The ACT-inconsistent measures received scores of 1, indicating that these
techniques were not used in treatment. In addition, although sessions were scored using the
standardized approach, the same reviewers were also trained to record observations at each
minute of the recording to identify which processes were being targeted in that time period.
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 13
Across all participants and therapy sessions, it was shown that each processes of change was
addressed the following percent of time: acceptance (30%), defusion (37%), self as context (4%),
being present (4%), values (10%), committed action (15%), and general assessment (23%).
PT
Cognitive challenging and stimulus management never occurred.
RI
Data Management
The data were collected using printed questionnaire materials, and the administrators of
SC
the questionnaires checked for missing data at the time of completion. There were no missing
NU
items. Three participants did not complete an assessment time point. As seen in Figure 1, one
participant in the ACT condition did not respond to attempts to schedule the follow-up
MA
assessment. One participant in the waitlist condition moved out of state during the waiting
period, and attempts to collect the remaining assessments were unsuccessful. Two participants in
D
TE
the waitlist condition received a partial intervention (six sessions and nine sessions) because they
moved out of state during the intervention. Of these two, one successfully completed the
P
remaining posttreatment and follow-up assessments. The other did not complete the remaining
CE
assessments. Last observation carried forward (LOCF) was not utilized in this study because
AC
moving posttreatment to follow-up for the ACT condition, and moving pretreatment to the end of
the waitlist for the waitlist condition, would have biased the outcomes in favor of the active
condition. Finally, LOCF was not used for the participant who completed 9 of 12 sessions but
did not complete post-assessment because this would not have accurately represented his gains
Data Analysis
Two sets of analyses were completed on the acquired data. First, “between condition”
analyses were conducted using the pretreatment and posttreatment data for all participants. These
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 14
MANOVA on the four measures. Because the interaction from the MANOVA was significant,
the univariate post hoc analyses were analyzed. Significant univariate 2X2 interactions were
PT
followed up with paired samples t-tests comparing pretreatment to posttreatment for each
RI
condition.
Because the waiting list condition received the treatment after the waiting period, both
SC
conditions’ data were combined and analyzed from pretreatment to posttreatment, and to follow-
NU
up. The “combined” analyses were completed by using the ACT condition’s pretreatment,
posttreatment, and three month follow-up data and the waitlist’s second assessment point
MA
(functioning as a pretreatment point because no treatment was delivered yet), posttreatment
(third assessment point), and three month follow-up (fourth assessment point). To ensure that the
D
TE
pretreatment assessment from each condition was equivalent, another randomization check was
conducted with select variables (variables that could have changed during the waiting period).
P
There were no significant differences between the immediate treatment condition and the second
CE
pretreatment assessment on measures of years of pornography use, average hours viewing per
AC
week, quality of life, or scores on the sexual compulsivity scale. These data were then combined
into a dataset with three timepoints and no group distinction: pretreatment, posttreatment, follow-
up. For the combined analyses a, 1-way (Time: pretreatment, posttreatment, follow-up)
MANOVA on the 4 measures was conducted. Because the multivariate effect was significant,
the univariate effects for time were interpreted for each measure. Pairwise post hoc comparisons
and follow-up were conducted for each measure following a significant effect for time.
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 15
Results
Between-condition Analyses
PT
showed significant multivariate effects for time F(4,22) = 5.35, p = .004, partial η2 = .5 and the
time X condition interaction F(4,22) = 3.17, p < .034, partial η2 = .366, but not for condition
RI
F(4,22) = 2.57, p < .066, partial η2 = .318. Follow-up univariate analyses showed that there were
SC
significant 2 (condition) X 2 (time) interactions for each dependent variable (ps ≤ .05). All
NU
means, standard deviations, and pretreatment to posttreatment effect sizes are reported in Table
2.
MA
For the DPVQ (self-reported hours per week of viewing pornography), univariate
analyses showed a significant condition X time interaction, F(1, 25) = 6.42, p = .018, partial η2 =
D
TE
.20. A post-hoc analysis comprised of paired samples t-tests (pretreatment to posttreatment) for
each condition showed a significant reduction in the ACT condition, t(13) = 4.81, p = .001, but
P
not the waitlist condition, t(12) = 1.21, p = .25. As shown in Table 2, the ACT condition
CE
For the QOLS, univariate analyses showed a significant condition X time interaction,
F(1, 25) = 4.23, p = .05, partial η2 = .15. A post-hoc analysis comprised of paired samples t-tests
for each condition did not show a significant reduction in the ACT condition, t(13) = -1.02, p =
.33, and the waitlist showed a significant worsening, t(12) = 2.37, p = .035.
For the SCS, univariate analyses showed a significant condition X time interaction, F(1,
25) = 7.8, p = .01, partial η2 = .24. A post-hoc analysis comprised of paired samples t-tests for
each condition showed a significant reduction in the ACT condition, t(13) = 4.76, p = .001, but
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 16
interaction, F (1, 25) = 10.28, p = .004, partial η2 = .291. A post-hoc analysis comprised of
PT
paired samples t-tests for each condition showed a significant reduction in the ACT condition,
RI
t(13) = 3.51, p = .004, but not the waitlist condition, t(12) = -.40, p = .7.
Combined Analyses
SC
A 1-way (Time: pretreatment, posttreatment, follow-up) MANOVA on the four measures
NU
showed a significant multivariate main effect for time F(8, 17) = 8.14, p <.001, partial η2 = .97.
Follow-up univariate analyses showed that there were significant effects for time on each
MA
measure. Table 3 shows means and standard deviations at the three time points, as well as
For the DPVQ, Mauchly’s test of sphericity was violated, χ2(2) = .76, p < .001; therefore,
TE
the Greenhouse-Geisser test is reported. Results showed a univariate significant main effect for
P
time, F(1.37, 24.89) = 28.38, p = .001, partial η2 = .55. Post-hoc analyses using a Bonferroni
CE
(p = .36).
For the QOLS, univariate analyses showed a significant main effect for time, F(2, 48) =
5.1, p = .01, partial η2 = .18. Posthoc analyses using a Bonferroni correction showed a significant
For the SCS, univariate analyses showed a significant main effect for time, F(2, 48) =
35.66, p < .001, partial η2 = .6. Post-hoc analyses using a Bonferroni correction showed a
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 17
For the CBOSB, univariate analyses showed a significant main effect for time, F(2, 48) =
PT
36.56, p = .001, partial η2 = .60. Post-hoc analyses using a Bonferroni correction showed a
RI
significant reduction from pretreatment to posttreatment (p = .001), and pretreatment to follow-
SC
Clinically Significant Change
NU
Because the present study is the first to examine a treatment for this problem, there are no
established clinical cutoff scores to evaluate clinical change. To present an indication of clinical
MA
significance from this investigation, the percentage of participants who obtained benchmark
participants (54%) had a 100% reduction in hours viewing, 17 (66%) had at least a 90%
reduction, 20 (78%) had at least an 80% reduction, and 24 (93%) had at least a 70% reduction. A
P
slight decline in response rates was seen from pretreatment to three month follow-up, where 9 of
CE
25 participants (36%) had a 100% reduction, 14 (56%) had at least a 90% reduction, 17 (68%)
AC
had at least an 80% reduction, and 19 (76%) had at least a 70% reduction.
Discussion
In this study, ACT for problematic pornography use was compared to a waitlist control
condition. The ACT condition showed a significant 93% decrease in hours viewed from
pretreatment to posttreatment compared to a 21% decrease in the control condition. After the
waitlist period, the control condition also received treatment, and results combined with the
original treatment condition showed a 92% reduction from pretreatment to posttreatment, and an
examining clinically significant change that showed that 54% of participants completely stopped
viewing at posttreatment, and that 93% of participants reduced viewing by at least 70% of
PT
participants had completely stopped viewing and 74% had reduced viewing by at least 70%.
RI
Additional analyses also showed that there was a 33% decrease on measures of sexual
behavior and cognitions. However, there was no evidence that ACT led to increased quality of
SC
life, calling into question its impact on broader functioning. It is possible that improvements in
NU
quality of life do not take place until reductions in pornography viewing have been present for
some time. Finally, independent evaluations showed that the treatment was implemented with
MA
integrity, and low treatment drop-out rates suggest that ACT for pornography viewing is
acceptable to participants.
D
TE
Even though problematic internet pornography use is not a diagnosable disorder, like
many behaviors that can become excessive, it can result in problems with functioning (Coleman
P
et al., 2001; McBride et al., 2008; Reid, 2007). It is also a clinical issue for which clinicians seek
CE
treatment guidance, but for which there are few empirically based suggestions (Del Giudice &
AC
Kutinsky, 2007; Reid & Woolley, 2006; Schneider, 1994; Young, 2007). At a theoretical level,
there are multiple studies showing that psychological inflexibility, the psychological process
addressed in ACT, predicts and mediates the severity of viewing (Levin et al., 2012; Twohig et
al., 2009). However, ours is the only randomized clinical trial addressing this issue. When
coupled with the previous study of ACT for problematic pornography use (Twohig & Crosby,
2010), the emerging picture is that this approach holds promise for helping individuals reduce
the nature of problematic viewing of pornography and the theory behind the treatment approach.
Most important is the role of experiential avoidance. As noted in the introduction, there is
PT
evidence to suggest that the way individuals interact with urges (i.e., thoughts, feelings, physical
RI
sensations) to view can result in increased distress and an increased rate of the behavior. In other
words, problematic behaviors are often worsened by a rigid and controlling reaction to the urges
SC
to act on the behavior. The core components of the ACT treatment approach were focused on
NU
changing the way that individuals react to the urges to view so that energy is directed at
managing the behavior, as opposed to managing the urges. The effectiveness of this approach
MA
provides some support that this theoretical approach to the problem is worth further
investigation.
D
TE
Future research should focus on ways to enhance the efficacy of this treatment. Given
that many of our participants were married or had significant others, it might be prudent to
P
involve those individuals in the treatment process. Treatment might also be enhanced by
CE
incorporating additional behavior therapy techniques that have proven to be effective with
AC
impulse control issues such as stimulus management (e.g., limiting physical computer access,
software controls that make viewing more difficult), habit reversal awareness and competing
response training, and more intensive self monitoring (Woods & Twohig, 2008). Maintenance
sessions might also help individuals maintain treatment gains. The ACT treatment protocol used
in this study was not augmented with these additional behavioral strategies to provide a better
There were several limitations to our study. First, as described in the participant
characteristics section, although our sample was heterogeneous with respect to age and marital
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 20
status, it was quite homogeneous with respect to sex, race, and geographical region. Religious
affiliation likely played an important role in participant motivation to seek and adhere to
treatment. Although not all individuals will be highly motivated to control their pornography
PT
viewing, there are many conservative religions that prohibit this behavior and many individuals
RI
have a moral stance against it. Second, all of our participants were male, thereby calling into
question the applicability of our findings to women. It is important to note, however, that
SC
pornography viewing tends to be more common among men (e.g., Twohig et al., 2009). Third,
NU
all outcomes were measured via self-report, which is subject to the effects of demand
characteristics, mood-memory effects, and other sources of inaccuracy. Future research should
MA
use objective measures of pornography viewing, such as computer monitoring programs, to
assess outcome. Fourth, our primary outcome variable was a self-report measure of hours spent
D
TE
viewing pornography, and this was not a formally validated measure with known psychometric
properties. Fifth, although this study attempted to experimentally control for threats to internal
P
validity, the waitlist design only controls for the passage of time. Thus, the superiority of the
CE
ACT condition could have been due to common factors of treatment, such as credibility and
AC
expectancy, increased attention to pornography viewing behavior, contact with a therapist, and
other nonspecific variables. There were also several methodological considerations that were not
included in this study, such as the use of an assessor who was unaware of treatment condition.
Finally, although an attempt was made to use multiple therapists, most of the treatment was
Overall, this study provides empirical evidence for an effective intervention for
problematic pornography use and highlights the need for continued work with this specific
problem, as well as other behavioral addictions. There is a clear need for more accurate
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 21
prevalence data, continued efforts to clarify the conceptualization of the problem, and further
References
PT
Baer, J. S., Stacy, A., & Larimer, M. (1991). Biases in the perception of drinking norms among
RI
college students. Journal of Studies on Alcohol, 52(6), 580-586.
SC
Black, D. W. (1998). Recognition and treatment of obsessive–compulsive spectrum disorders. In
NU
compulsive disorder: Theory, research, and treatment. (pp. 426-457). New York, NY
MA
US: Guilford Press.
Black, D. W., Kehrberg, L. L. D., Flumerfelt, D. L., & Schlosser, S. S. (1997). Characteristics of
D
154(2), 243-249.
P
Burckhardt, C. S., & Anderson, K. L. (2003). The Quality of Life Scale (QOLS): reliability,
CE
Coleman, E. (1991). Compulsive sexual behavior: New concepts and treatments. Journal of
AC
Coleman, E., Miner, M., Ohlerking, F., & Raymond, N. (2001). Compulsive Sexual Behavior
Inventory: A preliminary study of reliability and validity. Journal of Sex & Marital
Collins, R. L., Parks, G. A., & Marlatt, G. A. (1985). Social determinants of alcohol
consumption: The effects of social interaction and model status on the self-administration
10.1037/0022-006x.53.2.189
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 22
Cooper, A., Delmonico, D. L., & Burg, R. (2000). Cybersex users, abusers, and compulsives:
New findings and implications. Sexual Addiction & Compulsivity, 7(1-2), 5-29. doi:
10.1080/10720160008400205
PT
Cooper, A., Delmonico, D. L., Griffin-Shelley, E., & Mathy, R. M. (2004). Online sexual
RI
activity: An examination of potentially problematic behaviors. Sexual Addiction &
SC
Cooper, A., Griffin-Shelley, E., Delmonico, D. L., & Mathy, R. M. (2001). Online sexual
NU
problems: Assessment and predictive variables. Sexual Addiction & Compulsivity, 8(3-4),
Getting tangled in the net. Sexual Addiction & Compulsivity, 6(2), 79-104. doi:
D
TE
10.1080/10720169908400182
Dedmon, J. (2002). Is the Internet bad for your marriage? Online affairs, pornographic sites
P
Del Giudice, M. J., & Kutinsky, J. (2007). Applying motivational interviewing to the treatment
AC
of sexual compulsivity and addiction. Sexual Addiction & Compulsivity, 14(4), 303-319.
doi: 10.1080/10720160701710634
Dodge, B., Reece, M., Cole, S. L., & Sandfort, T. G. M. (2004). Sexual Compulsivity Among
10.1080/00224490409552241
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. (2012). Structured Clinical Interview
Grant, J. E., & Potenza, M. N. (2010). Impulse control disorders. In J. E. Grant & M. N. Potenza
(Eds.), Young adult mental health. (pp. 335-351). New York, NY US: Oxford University
Press.
PT
Hayes, S. C., Luoma, J. B., Bond, F. W., Masuda, A., & Lillis, J. (2006). Acceptance and
RI
commitment therapy: Model, processes and outcomes. Behaviour Research and Therapy,
SC
Hayes, S. C., Strosahl, K. D., & Wilson, K. G. (1999). Acceptance and commitment therapy: An
NU
experiential approach to behavior change. New York, NY US: Guilford Press.
Hilton Jr, D. L., & Watts, C. (2011). Pornography addiction: A neuroscience perspective.
MA
Surgical neurology international, 2, 19.
Kalichman, S. C., Johnson, J. R., Adair, V., Rompa, D., Multhauf, K., & Kelly, J. A. (1994).
D
TE
Sexual sensation seeking: Scale development and predicting AIDS-risk behavior among
10.1207/s15327752jpa6203_1
CE
Kalichman, S. C., & Rompa, D. (1995). Sexual sensation seeking and sexual compulsivity
AC
scales: Reliability, validity, and predicting HIV risk behavior. Journal of Personality
Kuzma, J. M., & Black, D. W. (2008). Epidemiology, prevalence, and natural history of
compulsive sexual behavior. Psychiatric Clinics of North America, 31(4), 603-611. doi:
10.1016/j.psc.2008.06.005
Levin, M. E., Hildebrandt, M. J., Lillis, J., & Hayes, S. C. (2012). The impact of treatment
10.1016/j.beth.2012.05.003
Levin, M. E., Lillis, J., & Hayes, S. C. (2012). When is online pornography viewing problematic
PT
among college males? Examining the moderating role of experiential avoidance. Sexual
RI
Addiction & Compulsivity, 19(3), 168-180.
Manning, J. C. (2006). The Impact of Internet Pornography on Marriage and the Family: A
SC
Review of the Research. Sexual Addiction & Compulsivity, 13(2-3), 131-165. doi:
NU
10.1080/10720160600870711
Marcks, B. A., & Woods, D. W. (2007). Role of thought-related beliefs and coping strategies in
MA
the escalation of intrusive thoughts: An analog to obsessive-compulsive disorder.
McBride, K. R., Reece, M., & Sanders, S. A. (2008). Predicting negative outcomes of sexuality
Mick, T. M., & Hollander, E. (2006). Impulsive-Compulsive Sexual Behavior. CNS Spectrums,
AC
11(12), 944-955.
Neighbors, C., Lee, C. M., Lewis, M. A., Fossos, N., & Larimer, M. E. (2007). Are social norms
Orzack, M. H., & Ross, C. J. (2000). Should virtual sex be treated like other sex addictions?
Owens, E. W., Behun, R. J., Manning, J. C., & Reid, R. C. (2012). The impact of internet
PT
Perry, M., Accordino, M. P., & Hewes, R. L. (2007). An investigation of Internet use, sexual and
RI
nonsexual sensation seeking, and sexual compulsivity among college students. Sexual
SC
Plumb, J. C., & Vilardaga, R. (2010). Assessing treatment integrity in acceptance and
NU
commitment therapy: Strategies and suggestions. International Journal of Behavioral
10.1080/10720160701480204
Reid, R. C., Bramen, J. E., Anderson, A., & Cohen, M. S. (2013). Mindfulness, Emotional
P
of clinical psychology.
AC
Reid, R. C., Carpenter, B. N., Hook, J. N., Garos, S., Manning, J. C., Gilliland, R., . . . Fong, T.
(2012). Report of findings in a DSM‐5 field trial for hypersexual disorder. Journal of
Reid, R. C., & Woolley, S. R. (2006). Using Emotionally Focused Therapy for Couples to
Rinehart, N. J., & McCabe, M. P. (1998). An empirical investigation of hypersexuality. Sexual &
Ross, M. W., Månsson, S.-A., & Daneback, K. (2012). Prevalence, severity, and correlates of
problematic sexual Internet use in Swedish men and women. Archives of Sexual
PT
Salisbury, R. M. (2008). Out of control sexual behaviours: A developing practice model. Sexual
RI
and Relationship Therapy, 23(2), 131-139. doi: 10.1080/14681990801910851
SC
diagnosis based on the DSM-III-R, and physician case histories. Sexual Addiction &
NU
Compulsivity, 1(1), 19-44. doi: 10.1080/10720169408400025
Schneider, J. P. (2000). Effects of cybersex addiction on the family: Results of a survey. Sexual
MA
Addiction & Compulsivity, 7(1-2), 31-58. doi: 10.1080/10720160008400206
Twohig, M. P., & Crosby, J. M. (2010). Acceptance and commitment therapy as a treatment for
D
TE
10.1016/j.beth.2009.06.002
P
Twohig, M. P., Crosby, J. M., & Cox, J. M. (2009). Viewing Internet pornography: For whom is
CE
it problematic, how, and why? Sexual Addiction & Compulsivity, 16(4), 253-266. doi:
AC
10.1080/10720160903300788
Winkler, A., Dörsing, B., Rief, W., Shen, Y., & Glombiewski, J. A. (2013). Treatment of internet
10.1016/j.cpr.2012.12.005
Young, K. S. (2007). Cognitive behavior therapy with Internet addicts: Treatment outcomes and
Author note:
This paper is based on Jesse Crosby’s dissertation at Utah State University. Dr. Crosby is now at
PT
Utah State University, [email protected].
RI
Table 1. ACT for Problematic Internet Pornography Use Treatment Components
SC
Session Treatment components Exercises and content
1 Informed consent Warning that therapy may result in emotional discomfort
NU
Commitment to complete all eight sessions
Limits to confidentiality Suicide, homicide, and abuse of children or disabled adults
The viewing of child pornography will be reported
MA
Values Increasing quality of life
Support client goals of either no viewing or reduced and controlled amounts of viewing
Acceptance Identify the distinction between viewing and urges to view
2 Acceptance Short-term vs. long-term effectiveness of attempts to control urges
D
metaphor
3 Acceptance Reinforce the futility of attempts to control urges
Identify attempts to control urges as part of the problem using the Polygraph, Chocolate
Cake, and What are the Numbers? exercises
P
Discussion of the social contexts that support regulation of private events using the Rule of
Private Events exercise
CE
PT
values using the Values Assessment Homework
Committed action Behavioral commitments to continue reduced viewing
Increased behavioral commitments to engage in valued living based on recent values work
RI
Discussion of relapse management using the ACT skills
11 Review Review any processes that still need attention
SC
12 Termination Summarize the treatment using the Joe the Bum metaphor
Apply ACT processes to relapse management
Apply ACT processes to termination
Suggest Get Out of Your Mind and Into Your Life workbook for continued progress
NU
Note. Italicized exercises are from Hayes and colleagues (1999).
MA
D
P TE
CE
AC
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 29
Table 2. Means, Standard Deviations, and Effect Sizes for Between Condition Analyses
PT
Pretreatment Posttreatment Pretreatment Posttreatment
───────── ──────── ───────── ─────────
Variable M SD M SD d M SD M SD d
RI
Self-report hours viewing 6.13 4.51 0.43 0.64 1.8 6.85 5.99 5.40 3.49 0.29
(per week)
SC
Quality of life (QOLS) 82.79 11.58 85.36 11.17 0.22 77.23 12.42 73.62 11.89 0.29
Sexual compulsivity (SCS) 30.43 8.36 20.36 7.27 1.28 31.15 7.36 30.08 6.87 0.07
Cognitive outcomes 38.14 5.19 30.64 5.30 1.43 39.77 5.42 40.23 5.66 -0.08
NU
(CBOSB)
Note: Cohen’s d was calculated using pooled standard deviation. All effect sizes are based on
MA
change scores from pretreatment.
D
P TE
CE
AC
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 30
Table 3. Means, Standard Deviations, and Effect Sizes for Combined Analyses
PT
Self-report hours viewing (per week) 5.65 4.01 0.47 0.80 1.79 0.77 1.10 1.66
Quality of life (QOLS) 78.37 12.42 85.77 9.98 0.65 84.32 12.06 0.48
RI
Sexual compulsivity (SCS) 30.26 7.54 19.15 6.44 1.58 19.72 7.65 1.38
Cognitive outcomes (CBOSB) 38.79 5.22 29.65 5.76 1.65 28.8 5.7 1.83
Note: Cohen’s d was calculated using pooled standard deviation. All effect sizes are based on
SC
change scores from pretreatment.
NU
MA
D
P TE
CE
AC
ACCEPTED MANUSCRIPT
ACT for Internet Pornography 31
PT
Commitment (n= 4) Assessed for Eligibility (n=39)
Excluded Total (n=11)
Did Not Meet Criteria (n=11)
Refusal to Participate (n=0)
Enrolled (n=28)
RI
SC
Pretreatment Assessment (n=28)
NU
ACT Condition (n=14)
MA Control (Waitlist) Condition (n=14)
HIGHLIGHTS
Acceptance and commitment therapy is more effective than waitlist in reducing viewing
Acceptance and commitment therapy had positive effects on related sexual behaviors
Results were maintained at three month follow-up
PT
RI
SC
NU
MA
D
P TE
CE
AC