Dialectical Behavior Therapy (DBT) Applied to College Students:
A Randomized Clinical Trial
Jacqueline Pistorello, Alan E. Fruzzetti, and
University of Nevada, Reno
West Chester University
Katherine M. Iverson
University of Nevada, Reno
Objective: College counseling centers (CCCs) are increasingly being called upon to treat highly
distressed students with complex clinical presentations. This study compared the effectiveness of
Dialectical Behavior Therapy (DBT) for suicidal college students with an optimized control
condition and analyzed baseline global functioning as a moderator. Method: The intent-to-treat
(ITT) sample included 63 college students between the ages of 18 and 25 years who were suicidal
at baseline, reported at least 1 lifetime nonsuicidal self-injurious (NSSI) act or suicide attempt, and
met 3 or more borderline personality disorder (BPD) diagnostic criteria. Participants were randomly
assigned to DBT (n ? 31) or an optimized treatment-as-usual (O-TAU) control condition (n ? 32).
Treatment was provided by trainees, supervised by experts in both treatments. Both treatments lasted
7–12 months and included both individual and group components. Assessments were conducted at
pretreatment, 3 months, 6 months, 9 months, 12 months, and 18 months (follow-up). Results: Mixed
effects analyses (ITT sample) revealed that DBT, compared with the control condition, showed significantly
greater decreases in suicidality, depression, number of NSSI events (if participant had self-injured), BPD
criteria, and psychotropic medication use and significantly greater improvements in social adjustment. Most
of these treatment effects were observed at follow-up. No treatment differences were found for treatment
dropout. Moderation analyses showed that DBT was particularly effective for suicidal students who were
lower functioning at pretreatment. Conclusions: DBT is an effective treatment for suicidal, multiproblem
college students. Future research should examine the implementation of DBT in CCCs in a stepped care
Keywords: DBT, college students, suicidality, borderline personality disorder
Although the college years are often thought of as carefree for
young adults, nearly half of this population can be diagnosed with
at least one mental health disorder in a given year (Blanco et al.,
2008). Depression, suicidal ideation, nonsuicidal self-injury
(NSSI), and borderline personality disorder (BPD) features are
significant mental health problems among college students. Ap-
This article was published Online First June 25, 2012.
Jacqueline Pistorello, Counseling Services, University of Nevada,
Reno; Alan E. Fruzzetti, Department of Psychology, University of Nevada,
Reno; Chelsea MacLane, Counseling Services, University of Nevada, Reno;
Robert Gallop, Department of Mathematics, Applied Statistics Program, West
Chester University; Katherine M. Iverson, Counseling Services and Psychol-
ogy Department, University of Nevada, Reno.
Chelsea MacLane is now at the Psychology Department, University of
Nevada, Reno. Katherine M. Iverson is now at the Women’s Health
Sciences Division of the National Center for PTSD, VA Boston Healthcare
System and Boston University School of Medicine.
The project described was supported by National Institute of Mental
Health Grant R34MH071904 to Jacqueline Pistorello. The authors would
like to thank Patricia Chatham in particular, as well as Grant Miller, Chad
Shenk, Victoria Follette, Jennifer Villatte, Karen Erickson, Larry Pruitt,
Sabrina Darrow, Susan Daflos, Melanie Watkins, Michael Katrichak,
Katrina Crenshaw, Lindsay Fletcher, and the faculty and staff at Counsel-
ing Services at the University of Nevada, Reno for their support. We also
want to thank Steven C. Hayes for statistical consultation and editing. Last,
we want to acknowledge the invaluable contribution made by the brave
young women and men who consented to participate in this research
The content is solely the responsibility of the authors and does not neces-
sarily represent the official views of the National Institute of Mental Health or
the National Institutes of Health. Pretreatment data from this study were
analyzed as part of a dissertation and appear in “An investigation of experi-
ential avoidance, emotion dysregulation, and distress intolerance in young
adult outpatients with borderline personality disorder symptoms,” by K. M.
Iverson, V. M. Follette, J. Pistorello, and A. E. Fruzzetti, in press, Personality
Disorders: Theory, Research, and Treatment. Data collected from therapists in
this study appear in two additional publications: “A preliminary examination
of burnout among counselor trainees treating clients with recent suicidal
ideation and borderline traits,” by G. D. Miller, K. M. Iverson, M. Kem-
melmeier, C. MacLane, J. Pistorello, A. E. Fruzzetti, . . . K. Y. Crenshaw,
2011, Counselor Education and Supervision, 50, pp. 344–359, and “A pilot
study of psychotherapist’s trainees’ alpha-amylase and cortisol levels during
treatment of recently suicidal clients with borderline traits,” by G. D. Miller,
K. M. Iverson, M. Kemmelmeier, C. MacLane, J. Pistorello, A. E. Fruzzetti,
. . . M. M. Watkins, 2010, Professional Psychology: Research and Practice,
41, pp. 228–235.
Correspondence concerning this article should be addressed to
Jacqueline Pistorello, Research and Clinical Faculty, Counseling Services,
University of Nevada, Reno, Mail Stop 0080, Reno, NV 89557. E-mail:
Journal of Consulting and Clinical Psychology
2012, Vol. 80, No. 6, 982–994
© 2012 American Psychological Association
proximately one third of college students report experiencing
depression that affected their ability to function in the past year
(American College Health Association, 2011). Suicide, a common
concomitant of depression, is a leading cause of death among
college students (Suicide Prevention Resource Center, 2004).
Nearly 20% of college students report seriously considering and
over 7% attempting suicide in their lifetime, whereas 6.4% of
students report contemplating and 1% attempting suicide in the
past year alone (American College Health Association, 2011).
Between 40% and 50% of severely suicidal students report mul-
tiple episodes of suicidal ideation (Drum, Brownson, Denmark, &
Smith, 2009). Nonsuicidal self-injury (NSSI), which refers to skin
cutting or other forms of intentional self-injury without conscious
suicidal intent, has been found to have a 15.3% lifetime and 6.8%
past-year prevalence rate among college students, with 43% of the
self-injurious students indicating six or more NSSI episodes
(Whitlock et al., 2011). Additionally, approximately 15% of intro-
ductory psychology students screen positive for significant BPD
features on a personality questionnaire (Trull, 1995), and 4% are
deemed to have a probable or definite BPD diagnosis in the
general campus population (Taylor, 2005).
Effective treatment of these problems among college students
may deflect the trajectory of these severe mental health problems
before they become chronic (Zivin, Eisenberg, Gollust, & Golber-
stein, 2009). College counseling centers (CCCs) are the front line
for providing mental health services for college students and are
increasingly involved in the treatment of severe psychopathology,
such as depression, suicidal ideation, and BPD (Benton, Robert-
son, Tseng, Newton, & Benton, 2003; Center for Collegiate Men-
tal Health, 2012), despite limited resources (Gallagher, 2011). The
implementation of empirically supported treatments for severe
and/or complex mental health problems in CCCs could have
significant and long-term public health impact; however, there is a
critical lack of studies on the effectiveness of such treatments in
these settings (Cooper, 2005; Kitzrow, 2003).
Dialectical behavior therapy (DBT; Linehan, 1993a, 1993b) is
an empirically supported treatment for individuals presenting with
complex and severe mental health problems, including BPD, de-
pression, suicidal ideation, and NSSI. DBT produces long-term
gains for suicidal BPD patients across a variety of domains,
including BPD symptoms, depression, suicidal ideation and at-
tempts, NSSI, and psychiatric hospitalization, and enhances social
functioning and global improvements (see Kliem, Kröger, & Ko-
sfelder, 2010, for a recent meta-analysis). Although several DBT
studies have focused on middle-aged women meeting full criteria
for BPD (e.g., Linehan, Armstrong, Suarez, Allmon, & Heard,
1991), there is evidence that DBT is feasible and effective with
male and female adolescents (Groves, Backer, van den Bosch, &
Miller, in press), as well as with other populations (Feigenbaum,
There are several reasons to investigate the implementation of
DBT in the treatment of college students with complex clinical
presentations, including suicidal ideation, severe depression,
NSSI, and BPD features. First, DBT is a principle-based treatment
that is flexible enough to apply to the severe and multiproblem
presentations increasingly seen across campuses (Benton et al.,
2003). Second, DBT focuses on teaching skills (e.g., emotion
regulation, distress tolerance) that are developmentally relevant to
college students (Kadison & DiGeronimo, 2004). Third, DBT was
designed for chronically suicidal individuals, and some experts
suggest that chronically suicidal students are more likely to benefit
from comprehensive treatment approaches and may actually ex-
perience iatrogenic effects with very brief forms of treatment
(Jobes, Jacoby, Cimbolic, & Husted, 1997). Fourth, the presenta-
tion of college students with BPD traits differs from community
BPD samples, and the treatment targets of DBT can be altered to
address college students’ specific clinical needs. For example,
college students are less likely than community BPD samples to
engage in recurrent suicidal threats (Trull, 1995), thus suggesting
that DBT treatment for this population may focus more on skills
acquisition than stabilization per se. Despite the logic behind its
application, there are no treatment outcome studies examining the
impact of DBT among college students.
Integrating DBT into CCCs needs to take into account that
CCCs are often designed to deliver short-term psychotherapy
(Center for Collegiate Mental Health, 2012), frequently rely on
trainees as therapists (Minami et al., 2009), and are subject to
relatively inflexible breaks and vacations, which can interfere with
manualized treatment approaches. Thus, research is needed to
evaluate whether DBT applied to the treatment of severely dis-
tressed college students within a CCC context is effective. Addi-
tionally, this research must examine whether treatment outcomes
are moderated by individual factors, such as students’ presenting
level of global functioning, in order to guide CCCs in terms of who
might best benefit from more intensive approaches. To this end,
the present study had two aims. The first aim was to compare DBT
to an optimized treatment-as-usual (O-TAU) condition in the treat-
ment of college students meeting three or more BPD diagnostic
criteria and reporting current suicidal ideation and a lifetime his-
tory of at least one NSSI and/or one suicide attempt at baseline.
The second aim was to analyze students’ baseline level of global
functioning as a moderator of treatment effects.
Participants were 63 college students seeking services at the
CCC at a medium-sized public university in the western United
States. Table 1 presents demographic characteristics of the sample.
Participants included in the study were between the ages of 18 and
25 years, reported suicidal ideation at baseline, as evidenced by a
score of 1 or higher on Question 9 of the Beck Depression
Inventory (2nd ed.; BDI-II; Beck, Steer, & Brown, 1996); en-
dorsed at least one act of lifetime NSSI and/or suicide attempt, as
measured by the Suicide Attempt-Self Injury Interview (SASII;
Linehan, Comtois, Brown, Heard, & Wagner, 2006); and met three
or more criteria on the BPD section of the Structured Clinical
Interview for DSM–IV Axis II Personality Disorders (SCID-II,
BPD; First, Spitzer, Gibbon, Williams, 1997). Participants were
excluded based on psychosis, need for inpatient care (as judged by
assessor), or prior DBT treatment and had to refrain from taking
part in other psychotherapy during the treatment portion of the
study. The local Institutional Review Board approved all study
Randomization and power.
signed to treatment conditions using a computerized urn random-
ization procedure (Stout, Wirtz, Carbonari, & Del Boca, 1994),
Participants were randomly as-
DBT APPLIED TO COLLEGE STUDENTS
controlling for gender, presence of NSSI or a suicide attempt
within the last 8 weeks, and psychotropic medication use at base-
line. Prior to conducting the study, a power analysis based on
preliminary findings from an ongoing DBT study relying on a
similar design (Linehan, Comtois, Murray, et al., 2006) and a DBT
study with a younger sample (Turner, 2000) showed that 31
participants per condition using an intent-to-treat (ITT) design
provided 80% power for d ? 0.8, with 15% attrition for primary
outcomes (e.g., suicide attempts and NSSI). These assumptions
were based on existing DBT studies that had obtained large effect
sizes for suicidal behaviors and very low attrition rates (e.g.,
Linehan, Armstrong, Suarez, Allmon, & Heard, 1991; Turner,
2000), and our original assumption of lower attrition among col-
Figure 1 depicts the participant flow through the study. Follow-
ing randomization to the two conditions, participants completed
assessments every 3 months during the treatment period (upper
limit of 12 months) and once again at follow-up (18 months after
the pretreatment assessment). Every effort was made to recruit
participants for assessments, including dropouts, and financial
incentives were offered for the 12- ($25) and 18-month ($75)
Axis-I Diagnostic Interview
Structured Clinical Interview for Axis I DSM–IV (Nonpatient
Version; psychotic screen, and Mood, Anxiety, Substance Use,
and Eating Disorders modules; SCID–I/NP; First, Spitzer, Gibbon,
& Williams, 2002) was used to gauge exclusion criteria and
describe the sample. This is a widely used instrument with accept-
able interrater reliability (First et al., 2002). Assessors were trained
until they reached 95% agreement on diagnostic categories.
Treatment Credibility Questionnaire
At the end of Session 1, participants rated the therapy they
were receiving on seven items adapted from Borkovec and Nau
(1972) that asked how logical, scientific, or potentially helpful
the treatment appeared to be. The coefficient alpha was .88 for
Outcome Measures: Interviews
dependent assessors, who were blind to treatment condition and
held masters or doctoral degrees in clinical psychology. Assessors
were trained by experts on the measures administered. A random
sample of 25% of the dependent variables’ videotapes were eval-
uated by an additional rater to calculate interrater reliability (kappa
for categorical measures, intraclass correlations [ICC] for ordinal
The SCID-II, BPD (First et al., 1997) served
as a screening and secondary outcome measure. It has good psy-
chometric properties and adequate convergent, discriminant, and
predictive validity (Ryder, Costa, & Babgby, 2007). This measure
was administered three times: at baseline, 12 months, and 18
months (follow-up), with the timeframe being “last year” for the
12-month assessment and “last 6 months” for the follow-up as-
Potential participants were interviewed by in-
Participant Characteristics at Pretreatment
?2, t test, or
Age (years), M (SD)
Living in a dorm (%)
Full-time student (%)
Born in the United States (%)
Year in school (%)
Current disorder (%)
Any major depressive disorder
Any substance-use disorder
Any anxiety disorder
Any eating disorder
B ? bisexual; T ? transgender; F. E. ? Fisher Exact test.
PISTORELLO, FRUZZETTI, MACLANE, GALLOP, AND IVERSON
sessment. Interrater reliability was good in this study, with ICCs
ranging from .85 to .93.
Suicide Attempt Self-Injury Interview (SASII; Linehan,
Comtois, Brown, et al., 2006).1
administered assessment of the frequency and topography of sui-
cide attempts and NSSI. ICCs for the subscales ranged from .87 to
.94 for this study. However, only the occurrence (yes/no) and
frequency of NSSI were analyzed in the present study, because the
base rate for NSSI and suicide attempts was too low post-baseline,
and the subscales can only be computed when NSSI or suicide
attempts are present.
The SASII is a clinician-
Outcome Measures: Self-Report Questionnaires
Beck Depression Inventory (2nd ed.; BDI-II; Beck et al.,
The BDI-II is a well-known measure of depressive symp-
tom severity and has good psychometric properties with this pop-
ulation (e.g., Steer & Clark, 1997). The coefficient alpha was .86
for this study.
Suicidal Behaviors Questionnaire (SBQ; Linehan, 1981).
The SBQ includes a scale measuring suicidality (SBQ-23), as well
as a table for self-report of NSSI and suicide attempts. On the SBQ
table, participants are asked to report the number of self-harming
acts across various categories (e.g., cutting, overdosing) and their
level of suicidal intent for each incident (e.g., clear intent to die,
ambivalent, no intent to die). Events were classified as suicide
attempts when the participant indicated ambivalent or clear intent
to die and as an NSSI when there was no intent to die. The SBQ-23
total score ranges from 0 to 88, assessing the frequency of suicidal
thoughts, and the person’s estimation of the likelihood they would
consider, attempt, and die from suicide in the future, across various
timeframes (e.g., next month, next 4 months, next year, lifetime).
The SBQ-23 has been used in other DBT studies (e.g., Linehan,
Comtois, Murray, et al., 2006), and its coefficient alpha was .92 for
Social Adjustment Scale-Self-Report (SAS-SR; Weissman &
The SAS-SR is a 54-item assessment of
functioning across a number of domains, which combine to
yield a total score. In this study, the total score yielded a
coefficient alpha of .81.
1For more information about measures created by Linehan and col-
leagues, see the Behavioral Research and Therapy Clinics website at
Participant flow chart. DBT ? dialectical behavior therapy; O-TAU ? optimized treatment as usual.
DBT APPLIED TO COLLEGE STUDENTS
Moderator: Independent Assessors’ Rating
Global Assessment of Functioning (GAF; Endicott, Spitzer,
Fleiss, & Cohen, 1976).
The GAF is a clinician-rated measure
of global functioning evaluated by assessors who were blind to
condition. Baseline GAF scores, based on a median split, were
used to test for moderation.
Overview of Treatments and Design
Participants were randomly assigned to either DBT or an
O-TAU condition, relying on supervision by an expert. Treat-
ment and data collection occurred between January 2006 and
The DBT treatment provided as part of this study
followed closely the standard outpatient DBT package (Linehan,
1993a, 1993b): (a) weekly 50-min individual psychotherapy (while
student was in town), (b) weekly 90-min group skills training, (c)
skills coaching as needed (via telephone, e-mail, or texting) be-
tween sessions to help patients generalize skills as solutions to
their difficulties (Linehan, 1993a), (d) weekly 90-min group su-
pervision/consultation for therapists, and (e) as-needed family
interventions (Fruzzetti, Santisteban, & Hoffman, 2007).
There were four modifications to standard DBT (Linehan,
1993a, 1993b) in this study. First, we shortened the distress toler-
ance module to 3 weeks and combined it with a 3-week-long
validation module utilized in recent DBT studies (e.g., Iverson,
Shenk, & Fruzzetti, 2009). Based on satisfaction ratings and exit
interviews from pilot work, college students who participated in
DBT group skills reported finding parts of the distress tolerance
module less helpful to them, possibly because college itself re-
quires a modicum of distress tolerance skills and the traditional
distress tolerance module was designed for a severe community
sample. Similarly, a validation module was rated as being very
helpful, perhaps because of the salience of interpersonal connec-
tions for college students, particularly those seeking treatment
(Center for Collegiate Mental Health, 2012). This validation mod-
ule included skills in recognizing different validation levels, val-
idating one’s own reactions, and validating others. Each skill area
(emotion regulation, interpersonal effectiveness, and distress tol-
erance/validation) lasted 6 weeks and was immediately preceded
by 2 weeks of mindfulness training, for a total of three modules,
each lasting 8 weeks.
Second, as is typical of DBT, therapists relentlessly encouraged
attendance and followed up on missed sessions through phone
calls and e-mails, but because students generally left town for
extended periods of time, participants had to miss four scheduled
consecutive individual appointments without contact with the ther-
apist to be considered a dropout. This may differ from typical
community DBT in that clients, even if away for 2–3 months
during scheduled university breaks, could still be considered to be
in ongoing treatment. Although the “four-miss rule” is usually
considered in the context of weekly contact, particularly until
client is stabilized, the DBT four-miss rule is, in fact, one way to
instantiate key DBT principles and targets, to maximize participa-
tion in treatment, including, (a) avoiding polarizing around missed
sessions (balancing the importance of attending with the realities
of some missed appointments) and (b) taking the therapist out of
the role of judge of what might be a “legitimate” versus “illegit-
imate” miss (instead, in DBT, when the client misses, we view it
as he or she is simply not present and does not benefit, regardless
of the reasons for missing; Linehan, 1993a). In this study, finding
a way to keep clients connected to treatment while recognizing
realistic constraints on physical attendance during university
breaks was the dialectical tension. We tried to resolve the tension
not only by not implementing a four-miss rule but also by using
nontraditional means to stay connected with clients. Depending on
participant severity, access, and motivation, long-distance sessions
via phone or Skype, and e-mail correspondence, were sometimes
used during academic breaks.
Third, skills groups ran for 1.5 hr to accommodate class sched-
ules. Fourth, the three 8-week skills modules followed the campus
schedule with one module taught in the spring, one in the fall, and
one in the summer (modules taught at each time varied depending
on the needs of enrolled clients). Optimal treatment in terms of
skills trainings involved attending eight group sessions per semes-
ter. Thus, only students completing 12 months of treatment re-
ceived all three skills training modules and some DBT completers
were only exposed to one or two skills group modules (e.g., may
have been gone in the summer or finished treatment early), al-
though skills training was also provided individually on occasion.
Students in this sample were only exposed to skills modules once,
whereas in traditional outpatient DBT that lasts 12 months clients
participate in each skills module at least twice (Linehan, 1993b).
We considered a treatment completer someone who stayed in
treatment between 7 and 12 months, regardless of number of
sessions attended. Although participants were offered up to 12
months of treatment, they were asked to make a commitment to
stay in treatment at least 7 months. The 7-month criterion was
selected to define completers because in pilot work many students
experienced significant improvement in one semester of treatment
and did not warrant continued intervention. The 7-month cutoff
(instead of only one semester) required students to attend treatment
across semesters, thus allowing counselors to ensure the student’s
therapeutic progress was stable. A shorter length of treatment has
been applied successfully in other DBT studies (e.g., Koons et al.,
Optimized treatment as usual (O-TAU).
Murray, et al. (2006) compared DBT with therapy by experts to
ensure high allegiance and expert treatment in the control condi-
tion. The present design emulated that approach, but because
treatment in CCCs is often provided by trainees (Minami et al.,
2009), we opted for a supervision by an expert paradigm. DBT
supervision was conducted by the first two authors, while a control
supervisor was selected who had been named by members of both
the Nevada State Board of Psychological Examiners and the Ne-
vada State Psychological Association as an expert in treating BPD
and suicidality using a non-cognitive–behavioral approach. The
O-TAU supervisor is the author of a psychodynamically oriented
book titled Treatment of Borderline Personality Disorder
(Chatham, 1985) that integrates developmental perspectives with
object relations theory.
The use of the supervision by experts design controls for the
supervisors’ allegiance and general expertise with this population
but does not lend itself to a rigorous comparison of treatments
(Linehan, Comtois, Murray, et al., 2006). The present comparison
reflects a well-trained and supervised (“optimized”) treatment
based on a coherent model—a realistic treatment-as-usual condi-
PISTORELLO, FRUZZETTI, MACLANE, GALLOP, AND IVERSON
tion. As such, it is not technically a comparison of two specific
treatments but, rather, one specific one (DBT) and one more
general (O-TAU). Thus, the O-TAU condition was not monitored
or coded for adherence. To attempt to more closely equate treat-
ment dose (contact time with treatment providers), O-TAU in-
cluded (a) once-weekly 50-min individual therapy (when student
was in town), (b) once-weekly 90-min group therapy (approxi-
mately 8 weeks per semester), (c) once-weekly 90-min group
supervision for therapists, (d) as-needed between-session consul-
tation, and (e) as-needed family interventions.
Trainee therapists and their training.
therapists in their respective conditions based on experience and
allegiance to the particular therapeutic orientation. Five DBT and
four O-TAU therapists were recruited for the study. Four of the
five DBT therapists and three of the four O-TAU therapists were
female. All DBT therapists were graduate students in a clinical
psychology program. One of the O-TAU psychotherapists was a
psychiatry resident, one was a psychology postdoctoral fellow, one
was a graduate student in clinical psychology, and one was a
master’s-level counseling psychology graduate student.
Trainees in both conditions underwent 30 hr of training in their
approach, separately (DBT or O-TAU), prior to beginning to offer
treatment. The DBT training followed Linehan (1993a, 1993b)
books and the O-TAU training was based on the text authored by
the O-TAU supervisor (Chatham, 1985). Interventions for both
conditions took place in the same CCC.
Treatment length and type.
were offered up to a year of treatment, and participants who
completed 7–12 months of treatment were considered completers.
Although both treatment conditions offered weekly individual and
group therapy, in DBT, skills groups are inherently part of the
treatment and attendance was emphasized, whereas in O-TAU, all
clients were offered group therapy but its emphasis was left up to
the clinical judgment of the therapist/supervisor.
Both conditions conducted 90-min weekly
group supervision. On average, each therapist maintained a case-
load of three clients and conducted one group. Additional individ-
ual supervision or phone consultation was provided as needed. In
DBT supervision, a brief update on all clients was provided with
more in-depth discussions prioritized based on DBT’s hierarchy
(e.g., suicidal and NSSI behaviors first). The structure and process
of the group supervision in O-TAU was left up to the supervisor.
Supervisors hired the
As noted above, participants
Analytic Approach for Treatment Outcome
Primary analyses were based on either mixed-model analysis of
variance (MMANOVA; Schwarz, 1993) or hierarchical linear
modeling (HLM; Raudenbush & Bryk, 2001), which take into
account missing data and the spacing between assessments (Ver-
beke & Mloenberg, 2000). HLM was used if the change appeared
linear; otherwise, an MMANOVA approach was used. Appropri-
ate covariance structures were determined by comparison of the –2
restricted likelihood, Akaike information criterion, both uncor-
rected (AIC) and corrected (AICc).
Normality of all measures was ensured and transformations
applied if needed. Denominator degrees of freedom were estimated
with the Kenward-Roger approximation (Kenward & Roger,
1997). Effect sizes for the MMANOVA and HLM analyses were
based on overall F test statistics (Rosenthal & Rosnow, 1991) and
slope differences (Raudenbush & Liu, 2001), respectively, and
interpreted using the cutoffs suggested by Cohen (1992). Pattern-
mixture models were used to assess whether important estimates
were dependent on missing data patterns (Hedeker & Gibbons,
1997) or on total number of hours in treatment. To examine
moderation, GAF scores were converted into a dichotomous vari-
able using a median split, and the interaction of the effects of
treatment with this dichotomous variable was added to primary
outcome analyses. The Jacobson, Follette, and Revenstorf (1984)
method was used to analyze reliable change and clinical signifi-
cance of 12-month and follow-up findings for suicidality and
depression, taking into account whether there is normative data
available (Jacobson & Truax, 1991). All dropouts and other miss-
ing data were scored as unimproved and included in the denomi-
nator of the calculations, providing a full ITT analysis.
Pretreatment Differences and Characterization of
Randomization was successful with regard to balancing the
three covariate variables: (a) gender (22.6% male in DBT, 15.6%
in O-TAU, ?2? 0.49, p ? .48), (b) currently taking a psychotropic
medication (41.9% in DBT, 37.5% in O-TAU; ?2? 0.12, p ?
.72), and (c) presence of NSSI or a suicide attempt in the past 2
months (64.5% in DBT, 65.6% in O-TAU; ?2? 0.01, p ? .93).
Table 1 shows demographic and diagnostic variables at baseline.
Table 2 shows descriptive data for all dependent variables. At
baseline, participants presented with an average score on the
BDI-II in the severe range (M ? 32.63, SD ? 10.27) and approx-
imately two thirds (65%) had engaged in either NSSI or a suicide
attempt in the past 2 months. Pre-treatment differences on para-
metric measures were compared using t tests or chi-square/Fisher’s
exact test and revealed no significant differences between condi-
tions on any demographic, diagnostic, or pretreatment dependent
Treatment Credibility, Dropout, and Dosage
There were no differences on treatment credibility, DBT M ?
5.19, SD ? 0.95; O-TAU M ? 5.05, SD ? 0.93, t(47) ? –0.527,
p ? .10. There were also no differences on treatment dropout and
length in treatment: Approximately 35% of the DBT group (11/31)
and 47% of the O-TAU group (15/32) dropped out before com-
pleting 7 months of therapy (?2? 0.84, p ? .35), 19% of the DBT
group (6/31) and 12.5% of the O-TAU group (4/32) completed
7–11 months of therapy (?2? 0.55, p ? .45), and 45% of the DBT
group (14/31) and 41% of the O-TAU group (13/32) remained in
treatment for the entire 12 months (?2? 0.132, p ? .76). There
was also no difference in the attendance of individual sessions,
DBT M ? 24.97, SD ? 12.48; O-TAU M ? 20.70, SD ? 11.94,
t(60) ? –1.29, p ? .10; however, given the differences inherent to
these therapeutic approaches, DBT participants attended signifi-
cantly more group sessions, DBT M ? 12.48, SD ? 10.0; O-TAU
M ? 5.0, SD ? 6.3, t(60) ? –3.52 (60), p ? .01, d ? 0.90
(0.37–1.41). Treatment completers in both conditions attended a
significantly higher number of sessions than treatment dropouts, in
terms of both individual and group sessions (p ? .01). The average
DBT APPLIED TO COLLEGE STUDENTS
Means (and SDs) or Percent Frequencies for All Dependent Variables Across Assessment Points
Month 18 (Follow-up)
Nonsuicidal Self-Injury (SASII; Any Occurrence)
Nonsuicidal Self-Injury (SBQ; Any Occurrence)
Suicide Attempts (SASII; Any Occurrence)
Suicide Attempts (SBQ-23; Any Occurrence)
Borderline Criteria (SCID-BPD)
Social Adjustment (SAS-SR Total Score)
Psychotropic Medication (Any Usage)
O-TAU ? optimized treatment as usual; DBT ? dialectical behavior therapy; SBQ-23 ? Suicidal Behaviors Questionnaire (Linehan, 1981); BDI ? Beck Depression Inventory (2nd ed.; Beck,
Steer, & Brown, 1996); Suicide Attempt-Self Injury Interview (Linehan, Comtois, Brown, Heard, & Wagner, 2006); BPD ? borderline personality disorder; SCID ? Structured Clinical Interview for
DSM–IV Personality Disorders (First, Spitzer, Gibbon, Williams, 1997); SAS-SR ? Social Adjustment Scale-Self-Report (Weissman & Bothwell, 1976). Between-condition significance at each time
point relied on one-way analysis of covariance.
?p ? .05.
??p ? .01.
???p ? .001.
PISTORELLO, FRUZZETTI, MACLANE, GALLOP, AND IVERSON
DBT completer attended 34.00 (SD ? 10.46) individual and 17.00
(SD ? 9.38) skills group sessions, whereas the average O-TAU
completer attended 30.00 (SD ? 6.76) individual and 6.70 (SD ?
6.78) group sessions. Therefore, the same pattern emerged among
completers: There was no difference in number of individual
sessions (p ? .10), but there was a difference in number of groups
attended (p ? .01).
Adherence to DBT
An independent DBT adherence rater (trained to reliability)
reviewed a random sample of approximately 10% of the DBT
therapists’ taped individual sessions (81 of 773), randomly se-
lected across the beginning, middle, and end of treatment, using
the DBT Expert Rating Scale (Linehan, Lockard, Wagner, &
Tutek, 1996)—the predecessor to the adherence system being used
currently by Linehan and colleagues at the University of Wash-
ington. Raters were reliable with Linehan’s raters using this scale.
The ordinal scales range from 0.0 (nonadhering) to 5.0 (perfect
adherence and competence). Studies vary on cutoff scores used to
establish adherence in DBT: in some studies a score of 3.8 or
higher has been used to indicate minimal adherence (e.g., Koons et
al., 2001), while a score of 4.0 has been used in others (e.g.,
Linehan, Comtois, Murray, et al., 2006). In the present study, the
average rating for the Acceptance subscale was 3.87 (range ?
3.50–4.43, SD ? 0.23), the Change subscale was 3.92 (range ?
3.55–4.40, SD ? 0.22), and the Overall DBT Adherence Score
was 3.92 (range ? 3.57–4.45). Group sessions were not rated for
DBT adherence nor were the O-TAU sessions.
Table 2 provides means and standard deviations for all measures
at each assessment point. Three a priori primary outcomes were
examined: suicidality, depression, and NSSI. Results are shown
graphically in Figure 2.
An HLM analysis revealed that DBT participants
showed significantly greater reductions in suicidality, as measured
by the SBQ-23—frequency of suicidal thoughts, and the person’s
estimation of the likelihood they would consider, attempt, and die
from suicide in the future—than did O-TAU participants during
the treatment period, t(57) ? 2.02, p ? .049, d ? 0.53 (0.02–1.03).
Extending time through follow-up showed a similar effect, t(57) ?
2.36, p ? .022, d ? 0.63 (0.12–1.13). Application of the pattern
mixture model indicated no significant dependency of the treat-
ment effect on retention, t(59) ? 1.39, p ? .17, and total hours in
treatment, t(55) ? 1.12, p ? .26.
Reliable change required difference scores of 12.94 or more on
the SBQ-23; 31% of O-TAU participants improved and 16%
worsened by the 12-month assessment, 37% improved and 22%
worsened at follow-up; 32% of DBT participants improved and
0% worsened at 12 months, 48% improved and 0% worsened at
follow-up. At 12 months, the two conditions did not differ in net
gains (subtracting reliable deterioration from improvement) but
did so at follow-up, ?2(1) ? 7.80, p ? .006. This was primarily
due to greater deterioration in the O-TAU condition at 12 months
and follow-up (Fisher’s exact p ? .05). Focusing just on those with
reliable gains, clinical significance on the SBQ-23, due to the
absence of norms, relied on change of one SD or more as the
criterion (two SDs yielded a target score below zero). There was no
significant difference at 12 months (16% in O-TAU; 29% in DBT),
but there was a trend toward a significant difference at follow-up,
25% in O-TAU; 45% in DBT; ?2(1) ? 2.82, p ? .094, suggesting
that condition differences may have extended beyond merely
greater deterioration in O-TAU.
An MMANOVA analysis of the BDI-II scores
with toeplitz covariance structure revealed a significant effect for
time, F(4, 172) ? 8.16, p ? .0001, d ? 0.78, and for condition,
t(54) ? 2.78, p ? .008, d ? 0.76 (0.24–1.26). Condition contrasts
at fixed time points revealed that the two conditions did not differ
significantly (p ? .32) until after 6 months of treatment, when
significant differences in favor of DBT emerged, F(1, 168) ? 7.33,
p ? .008, d ? 0.70 (0.18–1.20). These condition differences
continued at 12 months, t(168) ? 2.95, p ? .004, d ? 0.74
(0.22–1.24), and follow up, t(168) ? 3.19, p ? .002, d ? 0.80
(0.28–1.30). The application of the pattern mixture model indi-
cated no significant dependency of the treatment effect on reten-
tion, t(52) ? 0.54, p ? .59, and total hours in treatment, t(52) ?
0.14, p ? .89.
Reliable changes based on study values required change scores
of 10.65 or more on the BDI-II. Approximately 47% of O-TAU
participants improved and 0% worsened, both at 12 months and at
follow up; 61% of DBT participants improved and 0% worsened at
12 months, 68% improved and 0% worsened at follow-up. The two
conditions did not differ in net gains of reliable change. For those
showing reliable improvement, using the normative data on the
BDI-II (Steer & Clark, 1997), a score of 20 was needed for
clinically significant change. DBT was significantly better in gen-
erating clinically significant improvement at 12 months (34% in
O-TAU; 61% in DBT), ?2(1) ? 4.57, p ? .04, odds ratio (OR) ?
3.02 (1.08–8.44), and showed a trend toward significance at
follow up (41% in O-TAU; 64% in DBT), ?2(1) ? 3.60, p ? .058,
OR ? 2.66 (0.99–7.36).
Two different measures were used to examine change
in NSSI: an interview conducted by a trained assessor (SASII) and
client self-reported frequencies (SBQ). Unlike community popu-
lations typically treated in DBT studies (e.g., Linehan et al., 1991),
our inclusion criteria did not specify recent and/or frequent NSSI.
Therefore, not everyone in our sample had engaged in NSSI, and
even among those who had recent NSSI, the lethality (assessed by
an independent assessor) was relatively low, with most clients
(70%) who reported recent self-injury at baseline receiving a
lethality rating of either 1 (very low) or 2 (low) on a 1–6 scale, with
higher scores indicating more severity. In fact, both of these NSSI
measures showed a preponderance of zeros (?50%) at each post-
baseline measure. Thus, we conducted a zero-inflated negative
binomial (ZINB) model where outcome measures were based on
average NSSI over the longitudinal period. There was a 44.4%
occurrence of NSSI for DBT and 43.3% for O-TAU (p ? .10). But
given an occurrence, the mean NSSI count was significantly lower
for DBT (1.50 ? 1.12) than for O-TAU (5.23 ? 8.47), t(57) ?
–2.11, p ? .04. Virtually identical findings were obtained with the
SBQ, t(56) ? –3.20, p ? .002, suggesting that a significant
intervention effect, in both measures, is being driven by a lower
NSSI count in DBT than O-TAU, when an occurrence happened.
Attempts were also measured via interview
and self-report. In both instances, self-harming behaviors that were
reported as either ambivalent or with clear intent to die were
DBT APPLIED TO COLLEGE STUDENTS
classified as suicide attempts (as opposed to NSSI). As illustrated
in Table 2, however, the post-baseline rate of suicide attempts was
too low to warrant separate outcome analyses.
two phases of change on the SCID-BPD: baseline to 12 months
and 12 months to follow-up. On average, there was a greater
reduction for DBT versus O-TAU during treatment (difference
estimate ? 0.18, SE ? 0.041), t(83) ? 4.51, p ? .0001, d ? 1.19
We performed a piecewise model with
(0.60–1.67), but not during the follow-up period (difference esti-
mate ? –0.08, SE ? 0.088), t(74.3) ? –1.01, p ? .31, d ? 0.27
The total score of the SAS-SR (see Fig-
ure 2) was analyzed using HLM. Results revealed more improve-
ment for DBT patients compared to those in the O-TAU condition
in social adjustment (difference estimate ? 0.019, SE ? 0.007),
t(107) ? 2.62, p ? .01, d ? 0.69 (0.15–1.16).
Psychotropic medication use.
chotomous variable of either no medication (0) or at least one
This analysis relied on a di-
Inventory (2nd ed.; Beck, Steer, & Brown, 1996); DBT ? dialectical behavior therapy; O-TAU ? optimized
treatment as usual; SBQ23 ? Suicidal Behaviors Questionnaire (Linehan, 1981); SASII ? Suicide Attempt-Self
Injury Interview (Linehan, Comtois, Brown, Heard, & Wagner, 2006); SCID-II ? Structured Clinical Interview
for DSM–IV Personality Disorders (First, Spitzer, Gibbon, Williams, 1997); SAS ? Social Adjustment Scale-
Self-Report (Weissman & Bothwell, 1976).
Results for dependent variables across assessment points in both conditions. BDI ? Beck Depression
PISTORELLO, FRUZZETTI, MACLANE, GALLOP, AND IVERSON
psychotropic medication (1), covarying pretreatment usage. Hier-
archical generalized linear modeling (HGLM), used to address the
binary form of the outcome, indicated that DBT clients used
significantly fewer psychotropic medications post-baseline (differ-
ence estimate ? 0.025, SE ? 0.009), t(60.1) ? 2.79, p ? .007, d ?
0.74 (see Figure 2).
There was good balance with respect to level of global func-
tioning across treatment assignment, where 67.7% of DBT partic-
ipants fell in the higher functioning range versus 65.5% of O-TAU
participants using a median split of pretreatment GAF scores. A
significant three-way interaction was found for the Treatment ?
Time ? Pretreatment GAF score on the primary dependent vari-
able—suicidality, t(51) ? 2.13, p ? .038. Contrasts between
slopes in the HLM analysis were used to explore this interaction.
For higher functioning participants there were no significant dif-
ferences between conditions at month 12 (difference estimate ?
0.58, SE ? 0.544), t(56.6) ? 1.07, p ? .29, d ? 0.28 (–0.23–
0.76), or at follow-up (difference estimate ? 0.75, SE ? 0.687),
t(53.7) ? 1.10, p ? .28, d ? 0.29 (–0.22–0.77). Among the lower
functioning participants, there was a nonsignificant trend toward a
difference in favor of DBT at month 12 (difference estimate ?
1.57, SE ? 0.81), t(62.7) ? 1.93, p ? .058, d ? 0.51 (–0.02–
0.99), and a significant difference at follow-up (difference esti-
mate ? 2.58, SE ? 1.05), t(57.7) ? 2.45, p ? .017, d ? 0.65
(0.11–1.12). With regard to depression, as measured by the BDI-II,
there was no statistically significant moderation effect of global
functioning at baseline, t(52) ? 0.33, p ? .75.
Despite underfunding, an emphasis on short-term treatment
(Gallagher, 2011), and a dependence on trainees as therapists
(Minami et al., 2009), CCCs are under a tremendous pressure to
provide services for severely distressed college students (Kitzrow,
2003). This expanded role is not just driven by client demand; it
has also been raised by the institutional and human costs of
tragedies such as the shootings in recent years at Virginia Tech and
Northern Illinois University, which in both instances involved
students thought to have been suicidal (Schwartz & Kay, 2009).
The present study shows that DBT can be adapted and imple-
mented successfully in CCCs to treat students presenting with a
complex, multiproblem, suicidal profile. DBT delivered by trainee
therapists with regular supervision successfully, and differentially,
impacted suicidality, depression, NSSI count, BPD symptoms,
psychotropic medication use, and social adjustment. DBT partic-
ipants made significant gains, with depression moving from a
pretreatment score in the severely depressed range to a follow-up
score in the minimal range of scores for the BDI-II. Moreover,
none of the DBT clients exhibited reliable worsening in suicidality
over the course of treatment, whereas this was not the case in the
O-TAU condition. This finding suggests that DBT may be a
particularly effective and safe treatment for severely distressed
clients being treated in the CCC context.
DBT adherence ratings indicate that, on average, the trainee
therapists in this study were sufficiently adherent to DBT treat-
ment strategies and interventions. Because we endeavored to
mimic “real world” training and supervision conditions, we did not
require therapists to achieve adherence levels prior to taking on
research treatment cases. Thus, these mean adherence scores re-
flect both early sessions, with generally lower adherence scores,
and later sessions, with generally higher adherence scores. Overall,
mean scores around 3.9 suggest reasonably high levels of learning
and adherence and suggest that these can be achieved through
moderate levels of training and supervision that could be replicated
in other counseling centers.
These positive findings in favor of DBT are unlikely to be due
to methodological weaknesses such as a weak control condition,
higher expectations by experimental participants, lack of expertise
by clinician/supervisor in treating BPD, differential access to treat-
ment, differential training, or differential allegiance to treatment.
Both treatments were provided within the same clinic, overseen by
expert supervisors with strong allegiance to their particular treat-
ment approach, and delivered by therapists with similar training
and supervision intensity.
The O-TAU comparison condition appears to have been effec-
tive in its own right. The supervisor was selected based on com-
munity recognition of expertise with this population (similar to
Linehan, Comtois, Murray, et al., 2006), and the O-TAU condition
showed significant improvement (within-subject) on most mea-
sures. In depression, for example, the impact of O-TAU and DBT
was comparable until about 6 months of treatment (see Figure 2),
and average endpoint depression levels for O-TAU were within
one standard deviation of the clinical cutoff of the BDI-II. How-
ever, this control condition was optimized. Therapists in this
condition were trained and supervised by an expert in suicidality
and BPD, who published in this area and had extensive experience
supervising trainees with difficult cases. This finding may not
generalize to treatment provided by CCC trainees not supervised
by suicidality experts.
This study surprisingly obtained higher dropout rates (35%)
than some other DBT studies (e.g., Linehan et al., 1991). There are
a few possible explanations. First, the natural breaks in treatment
at a CCC (e.g., winter and summer breaks) may have interrupted
treatment enough that it was difficult for participants to reconnect.
Second, the sample recruited for this study was, on average,
somewhat higher functioning than community DBT studies, with
participants not being as likely to be engaging in severe NSSI or
suicide attempts, only meeting three or more BPD criteria, and
being able to function, at least at baseline, in a college setting.
Thus, unlike community samples, these college students may have
been able to maintain some functionality in the short-term, even
when not in treatment. Last, but not least, not all students stopping
treatment before 7 months of treatment were, in fact, clinical
dropouts: Some students across both conditions mentioned at the
time or later during exit interviews that they felt better and no
longer needed the treatment, and some were students who trans-
ferred to different schools and were simply unable to continue. We
did foresee that there might be a subsegment of the student
population who might recover more rapidly than others—hence
our decision to allow participants to terminate therapy within a
range of time (7–12 months), as opposed to requiring every student
to stay in a comprehensive approach such as DBT for the full year.
In the future, the association between treatment dosage to out-
come in DBT should be studied systematically, including random-
ized trials where participants are assigned to different lengths
DBT APPLIED TO COLLEGE STUDENTS
and/or modalities of DBT treatment. This study’s higher dropout
rate and participants’ reduced access to skills groups due to cam-
pus schedules has resulted, on average, in relatively fewer number
of DBT sessions than that expected for weekly treatments lasting
up to 1 year (i.e., the mean number of individual sessions among
DBT treatment completers was 34, SD ? 10). Thus, although the
findings favor DBT, it is possible that the limited number of
individual and group sessions completed could have reduced the
potency of the DBT treatment.
Moderation analyses suggest that DBT might be particularly
effective for complex suicidality among college students who are
lower in global functioning (GAF baseline score 50 and below).
This is not particularly surprising, as DBT was originally devel-
oped to work with lower functioning individuals (Linehan et al.,
1991); however, this finding has important implications for match-
ing clients to treatments at CCCs. These findings suggest that a
stepped care approach might be viable in which the full outpatient
DBT package may only be delivered to those with a lower func-
tioning profile (i.e., college students who present with severe
suicidality and depression, BPD features, and lower GAF scores).
There may be other opportunities for efficiency that should be
explored in research with severe cases in CCCs.
Despite multiple strengths, this study has methodological limi-
tations. It is not possible to conclude from the present study that DBT
outperformed psychodynamically oriented therapy, despite the fact
that the O-TAU supervisor is well known in this area, because the
implementation of the O-TAU intervention, as a psychodynamic
approach, was not carefully controlled or monitored for adherence.
Moreover, O-TAU and DBT therapists differed somewhat in terms of
differences may have affected results.
Although total number of hours in treatment did not moderate
treatment effects, there were significant treatment differences in
attendance of group sessions, which emerged as an inherent dif-
ference between these approaches. DBT requires clients to partic-
ipate in both the individual and group components of therapy to
remain in treatment. Beyond requiring that O-TAU offer both
individual and group treatment (weekly), the study did not impose
conditions on how treatment should be carried out, as more intru-
sive requirements could have fundamentally changed the O-TAU
treatment, resulting in essentially developing an ad hoc manualized
alternative, which was not the intent of the study design.
Similarly, the present study speaks to the effectiveness of a
comprehensive DBT package (individual, group, between-session
consultation to the patient, and consultation team for therapists) in
treating students presenting to CCCs but cannot address the role of
DBT components if applied separately. Most CCC administrators
might prefer to offer DBT skills groups only, to keep costs low.
However, in comprehensive DBT, the individual therapist plays a
major role in promoting group attendance and motivating the
student to get back into group after missing session, and ensuring
the generalization of skills and having a consultation team for
therapists are also essential (Linehan, 1993a). Whether DBT group
skills training alone can be effective with a severe multiproblem
college sample will need to be addressed in future studies. More-
over, this study was powered to ask outcome questions, and
although simple moderation analyses were successfully performed
analyzing Treatment Outcome ? Level of Student Global Func-
tioning, the study was underpowered to address other possible
interactions, such as between level of global functioning and
participants’ diagnostic and clinical profiles (e.g., comorbid sub-
stance abuse) or demographic characteristics (e.g., gender, ethnic-
There are noteworthy limitations in terms of adherence proce-
dures and the assessment timeline. Although we utilized the DBT
adherence system available when this project was first proposed, it
is not the most current one, and although purposeful for other
reasons, our study did not require DBT therapists to achieve
adherence levels before starting to see clients—a factor that may
have influenced mean adherence ratings. Additionally, O-TAU
sessions were not rated for DBT adherence, and therefore, the
study failed to address treatment diffusion. Last, this study was
envisioned as a combination of effectiveness and efficacy ele-
ments, requiring some compromise relative to an efficacy-only
randomized trial. For example, allowing participants to stay in
treatment between 7 and 12 months meant that not all participants
completed a post-test treatment assessment at the same time. The
use of mixed effects analyses in evaluating treatment effects
helped compensate for this design feature as they detect different
change patterns across condition and time, instead of relying on
differences at specific time points.
Universities are increasingly faced with difficult choices of how to
allocate resources, but it seems clear that CCCs must find a way to
address complex and severe problems of students on campus (Gal-
lagher, 2011). The increase in severity of student problems, and the
large institutional and human costs (e.g., lawsuits, publicity, concerns
about campus safety) that come from failing to prevent or to treat
severe problems means that they cannot be ignored (Schwartz & Kay,
2009). This is a change in the historic role of CCCs (Kitzrow, 2003)
and the research and treatment development community need to help
CCCs find an effective new posture.
In that context, evidence-based treatments in general offer hope
to students and to the institution alike (Cooper, 2005). The tran-
sition into adulthood is a particularly unique period and is a time
when individuals are likely to experience the onset of mental
disorders that might persist (Zivin et al., 2009). Colleges and
universities directly touch the lives of nearly half of the population
of the United States (Stoops, 2004). If DBT can effectively treat
severe mental health problems, such as suicidality, depression, and
NSSI, among college students, the public health of the entire
population could be impacted over time.
American College Health Association. (2011). ACHA-National College
Health Assessment II: Reference group executive summary Spring 2011.
Hanover, MD: Author.
Beck, A. T., Steer, R. A., & Brown, K. G. (1996). Manual for the Beck
Depression Inventory-II. San Antonio, TX: Psychological Corporation.
Benton, S. A., Robertson, J. M., Tseng, W., Newton, F. B., & Benton, S. L.
(2003). Changes in counseling center client problems across 13 years.
Professional Psychology, Research and Practice, 34, 66–72. doi:
Blanco, C., Okuda, M., Wright, C., Hasin, D. S., Grant, B. F., Liu, S. M.,
& Olfson, M. (2008). Mental health of college students and their
non-college-attending peers: Results from the national epidemiologic
study on alcohol and related conditions. Archives of General Psychiatry,
65, 1429–1437. doi:10.1001/archpsyc.65.12.1429
Borkovec, T. D., & Nau, S. D. (1972). Credibility of analogue therapy
PISTORELLO, FRUZZETTI, MACLANE, GALLOP, AND IVERSON
rationales. Journal of Behavior Therapy and Experimental Psychiatry, 3,
Center for Collegiate Mental Health. (2012). 2011 annual report (Publi-
cation No. STA 12-59). Retrieved from http://ccmh.squarespace.com
Chatham, P. M. (1985). Treatment of the borderline personality. New
York, NY: Jason Aronson.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.
Cooper, S. E. (2005). Evidence-based psychotherapy practice in college
mental health. Journal of College Student Psychotherapy, 20, 1–6.
Drum, D. J., Brownson, C., Denmark, A., & Smith, S. E. (2009). New data
on the nature of suicidal crises in college students: Shifting the para-
digm. Professional Psychology, Research and Practice, 40, 213–222.
Endicott, J., Spitzer, R. L., Fleiss, J. L., & Cohen, J. (1976). The global
assessment scale: A procedure for measuring overall severity psychiatric
disturbance. Archives of General Psychiatry, 33, 766–771. doi:10.1001/
Feigenbaum, J. (2007). Dialectical behaviour therapy: An increasing evi-
dence base. Journal of Mental Health, 16, 51–68. doi:10.1080/
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (1997).
Structured Clinical Interview for DSM–IV Personality Disorders (SCID-II).
Washington, DC: American Psychiatric Press.
First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (2002).
Structured Clinical Interview for DSM–IV Axis I Disorders, Research
Version, Non-Patient Edition (SCID-I/NP). New York, NY: Biometrics
Research, New York State Psychiatric Institute.
Fruzzetti, A. E., Santisteban, D. A., & Hoffman, P. D. (2007). Dialectical
behavior therapy with families. In L. A. Dimeff & K. Koerner (Eds.),
Dialectical behavior therapy in clinical practice: Applications across
disorders and settings (pp. 222–244). New York, NY: Guilford Press.
Gallagher, R. P. (2011). National survey of counseling center directors.
Alexandria, VA: International Association of Counseling Services.
Groves, S. S., Backer, H. S., van den Bosch, L. M. C., & Miller, A. L. (in
press). Dialectical behavior therapy with adolescents: A review. Child
and Adolescent Mental Health.
Hedeker, D., & Gibbons, R. D. (1997). Application of random-effects
pattern-mixture models for missing data in longitudinal studies. Psycho-
logical Methods, 2, 64–78. doi:10.1037/1082-989X.2.1.64
Iverson, K., Shenk, C., & Fruzzetti, A. E. (2009). Dialectical behavior therapy
for women victims of domestic abuse: A pilot study. Professional Psychol-
ogy: Research and Practice, 40, 242–248. doi:10.1037/a0013476
Jacobson, N. S., Follette, W. C., & Revenstorf, D. (1984). Psychotherapy
outcome research: Methods for reporting variability and evaluating
clinical significance. Behavior Therapy, 15, 336–352. doi:10.1016/
Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical
approach to defining meaningful change in psychotherapy research.
Journal of Consulting and Clinical Psychology, 59, 12–19. doi:10.1037/
Jobes, D. A., Jacoby, A. M., Cimbolic, P., & Hustead, L. A. T. (1997).
Assessment and treatment of suicidal clients in a university counseling
center. Journal of Counseling Psychology, 44, 368–377. doi:10.1037/
Kadison, R., & DiGeronimo, T. F. (2004). College of the overwhelmed:
The campus mental health crisis and what to do about it. San Francisco,
Kenward, M. G., & Roger, J. H. (1997). Small sample inference for fixed
effects from restricted maximum likelihood. Biometrics, 53, 983–997.
Kitzrow, M. A. (2003). The mental health needs of today’s college stu-
dents: Challenges and recommendations. The National Association of
Student Personnel Administrators Journal, 41, 167–181.
Kliem, S., Kröger, C., & Kosfelder, J. (2010). Dialectical behavior therapy
for borderline personality disorder: A meta-analysis using mixed effects
modeling. Journal of Consulting and Clinical Psychology, 78, 936–951.
Koons, C. R., Robins, C. J., Tweed, J. L., Lynch, T. R., Gonzalez, A. M.,
Morse, J. Q., . . . Bastian, L. A. (2001). Efficacy of dialectical behavior
therapy in women veterans with borderline personality disorder. Behav-
ior Therapy, 32, 371–390. doi:10.1016/S0005-7894(01)80009-5
Linehan, M. M. (1981). Suicidal behaviors questionnaire. Unpublished
manuscript, University of Washington, Seattle, WA.
Linehan, M. M. (1993a). Cognitive behavioral therapy of borderline per-
sonality disorder. New York, NY: Guilford Press.
Linehan, M. M. (1993b). Skills training manual for borderline personality
disorder. New York, NY: Guilford Press.
Linehan, M. M., Armstrong, H. E., Suarez, A., Allmon, D., & Heard, H. L.
(1991). Cognitive behavioral treatment of chronically parasuicidal bor-
derline patients. Archives of General Psychiatry, 48, 1060–1064. doi:
Linehan, M. M., Comtois, K. A., Brown, M. Z., Heard, H. L., & Wagner,
A. (2006). Suicide Attempt Self-Injury Interview (SASII): Development,
reliability, and validity of a scale to assess suicide attempts and inten-
tional self-injury. Psychological Assessment, 18, 303–312. doi:10.1037/
Linehan, M. M., Comtois, K. A., Murray, A. M., Brown, M. Z., Gallop, R. J.,
Heard, H. L., . . . Lindenboim, N. (2006). Two-year randomized controlled
trial and follow-up of dialectical behavior therapy vs. therapy by experts for
suicidal behaviors and borderline personality disorder. Archives of General
Psychiatry, 63, 757–766. doi:10.1001/archpsyc.63.7.757
Linehan, M. M., Lockard, J. S., Wagner, A. W., & Tutek, D. (1996). DBT
Expert Rating Scale. Unpublished manuscript, University of Washing-
ton, Seattle, WA.
Minami, T., Davies, D. R., Callan Tierney, S., Bettman, J. E., McAward,
S. M., Averill, L. A, . . . Wampold, B. E. (2009). Preliminary evidence
on the effectiveness of psychological treatments delivered at a university
counseling center. Journal of Counseling Psychology, 56, 309–320.
Raudenbush, S., & Bryk, A. (2001). Hierarchical linear modeling: Appli-
cations and data analysis methods. Newbury Park, CA: Sage.
Raudenbush, S. W., & Liu, X. (2001). Effects of study duration, frequency
of observation, and sample size on power in studies of group differences
in polynomial change. Psychological Methods, 6, 387–401. doi:
Rosenthal, R., & Rosnow, R. L. (1991). Essentials of behavioral research,
methods, and data analysis. San Francisco, CA: McGraw-Hill.
Ryder, A. G., Costa, P. T., & Bagby, R. (2007). Evaluation of the SCID-II
personality disorder traits for DSM–IV: Coherence, discrimination, re-
lations with general personality traits, and functional impairment. Jour-
nal of Personality Disorders, 21, 626–637. doi:10.1521/pedi.2007
Schwartz, V., & Kay, J. (2009). The crisis in college and university mental
health. Psychiatric Times, 26, 32–34.
Schwarz, C. J. (1993). The mixed-model ANOVA: The truth, the computer
packages, the books. American Statistician, 47, 48–59.
Steer, R. A., & Clark, D. A. (1997). Psychometric characteristics of the
Beck Depression Inventory–II with college students. Measurement and
Evaluation in Counseling and Development, 30, 128–136.
Stoops, N. (2004). Educational attainment in the United States: 2003.
Population characteristics (Publication No. P20–550). Washington,
DC: U.S. Census Bureau.
Stout, R. L., Wirtz, P. W., Carbonari, J. P., & Del Bocca, F. K. (1994).
Ensuring balance distribution of prognostic factors in treatment outcome
research. Journal of Studies on Alcohol, 12, 70–75.
DBT APPLIED TO COLLEGE STUDENTS
Suicide Prevention Resource Center. (2004). Promoting mental health and Download full-text
preventing suicide in college and university settings. Newton, MA:
Education Development Center.
Taylor, J. (2005). Substance use disorders and Cluster B personality
disorders: Physiological, cognitive, and environmental correlates in a
college sample. American Journal of Drug and Alcohol Abuse, 31,
Trull, T. J. (1995). Borderline personality disorder features in nonclinical
young adults: Identification and validation. Psychological Assessment,
7, 33–41. doi:10.1037/1040-35126.96.36.199
Turner, R. M. (2000). Naturalistic evaluation of dialectical behavior
therapy-oriented treatment for borderline personality disorder. Cognitive
and Behavioral Practice, 7, 413–419. doi:10.1016/S1077-7229(00)
Verbeke, G., & Molenberghs, G. (2000). Linear mixed models for longi-
tudinal data. New York, NY: Spinger-Verlag.
Weissman, M. M., & Bothwell, S. (1976). Assessment of social adjustment
by patient self-report. Archives of General Psychiatry, 33, 1111–1115.
Whitlock, J, Muehlenkamp, J., Purington, A., Eckenrode, J., Barreira, J.,
Abrams, G. B., . . . Knox, K. (2011). Non-suicidal self-injury in a
college population: General trends and sex differences. Journal of Amer-
ican College Health, 58, 691–698.
Zivin, K., Eisenberg, D., Gollust, S. E., & Golberstein, E. (2009). Persis-
tence of mental health problems and needs in a college student popula-
tion. Journal of Affective Disorders, 117, 180–185. doi:10.1016/
Received October 10, 2011
Revision received May 10, 2012
Accepted May 14, 2012 ?
Call for Nominations
The Publications and Communications (P&C) Board of the American Psychological Association
has opened nominations for the editorships of Behavioral Neuroscience, Journal of Applied
Psychology, Journal of Educational Psychology, Journal of Personality and Social Psychology:
Interpersonal Relations and Group Processes, Psychological Bulletin, and Psychology of
Addictive Behaviors for the years 2015–2020. Mark S. Blumberg, PhD, Steve W. J. Kozlowski,
PhD, Arthur Graesser, PhD, Jeffry A. Simpson, PhD, Stephen P. Hinshaw, PhD, and Stephen A.
Maisto, PhD, ABPP, respectively, are the incumbent editors.
Candidates should be members of APA and should be available to start receiving manuscripts in
early 2014 to prepare for issues published in 2015. Please note that the P&C Board encourages
participation by members of underrepresented groups in the publication process and would partic-
ularly welcome such nominees. Self-nominations are also encouraged.
Search chairs have been appointed as follows:
● Behavioral Neuroscience, John Disterhoft, PhD
● Journal of Applied Psychology, Neal Schmitt, PhD
● Journal of Educational Psychology, Neal Schmitt, PhD, and Jennifer Crocker, PhD
● Journal of Personality and Social Psychology: Interpersonal Relations and Group Pro-
cesses, David Dunning, PhD
● Psychological Bulletin, Norman Abeles, PhD
● Psychology of Addictive Behaviors, Jennifer Crocker, PhD, and Lillian Comas-Diaz, PhD
Candidates should be nominated by accessing APA’s EditorQuest site on the Web. Using your
Web browser, go to http://editorquest.apa.org. On the Home menu on the left, find “Guests.” Next,
click on the link “Submit a Nomination,” enter your nominee’s information, and click “Submit.”
Prepared statements of one page or less in support of a nominee can also be submitted by e-mail
to Sarah Wiederkehr, P&C Board Search Liaison, at email@example.com.
Deadline for accepting nominations is January 11, 2013, when reviews will begin.
PISTORELLO, FRUZZETTI, MACLANE, GALLOP, AND IVERSON