Journal of Affective Disorders 200 (2016) 51–57

Contents lists available at ScienceDirect

Journal of Affective Disorders

journal homepage: www.elsevier.com/locate/jad

Research paper

The Effects of Mindfulness‐Based Cognitive Therapy and Cognitive

Behavioral Analysis System of Psychotherapy added to Treatment as
Usual on suicidal ideation in chronic depression: Results of a
randomized-clinical trial
Thomas Forkmann a,n, Eva-Lotta Brakemeier b, Tobias Teismann c, Elisabeth Schramm d,
Johannes Michalak e
a
Institute of Medical Psychology and Medical Sociology, University Hospital of RWTH Aachen University, Germany
b
Department of Clinical Psychology and Psychotherapy, Berlin University of Psychology, Germany
c
Department of Clinical Psychology and Psychotherapy, Ruhr-Universität Bochum, Germany
d
Department of Psychiatry and Psychotherapy, University of Freiburg, Germany
e
Department of Psychology and Psychotherapy, Witten/Herdecke University, Germany

art ic l e i nf o a b s t r a c t

Article history: Background: Suicidal ideation (SI) is common in chronic depression, but only limited evidence exists for
Received 8 October 2015 the assumption that psychological treatments for depression are effective for reducing SI.
Received in revised form Methods: In the present study, the effects of Mindfulness-based Cognitive Therapy (MBCT; group ver-
5 January 2016
sion) plus treatment-as-usual (TAU: individual treatment by either a psychiatrist or a licensed psy-
Accepted 24 January 2016
Available online 19 April 2016
chotherapist, including medication when indicated) and Cognitive Behavioral Analysis System of Psy-
chotherapy (CBASP; group version) plus TAU on SI was compared to TAU alone in a prospective, bi-center,
Keywords: randomized controlled trial. The sample consisted of 106 outpatients with chronic depression.
Mindfulness-based Cognitive Therapy Results: Multivariate regression analyses revealed different results, depending on whether SI was as-
Cognitive Behavioral Analysis System of
sessed via self-report (Beck Depression Inventory suicide item) or via clinician rating (Hamilton De-
Psychotherapy
pression Rating Scale suicide item). Whereas significant reduction of SI emerged when assessed via
Suicidal ideation
Chronic depression clinician rating in the MBCT and CBASP group, but not in the TAU group while controlling for changes in
Persistent depression depression, there was no significant effect of treatment on SI when assessed via self-report.
RCT Limitations: SI was measured with only two single items.
Conclusions: Because all effects were of small to medium size and were independent of effects from other
depression symptoms, the present results warrant the application of such psychotherapeutical treatment
strategies like MBCT and CBASP for SI in patients with chronic depression.
& 2016 Elsevier B.V. All rights reserved.

1. Introduction et al., 2003). Suicidal ideation (SI) in chronic depression tends to

Suicide is among the 10 leading causes of death in most
countries (WHO, 2014) and the number of suicide attempts is
estimated to be up to fifteen times higher than the number of
actual deaths by suicide (Borges et al., 2010). About 40% of in-
dividuals who died by suicide experienced depressive illness be-
fore their death (Arsenault-Lapierre et al., 2004). Chronic depres-
sion, which represents approximately 20–30% of all depressive
disorders, is known to have a higher risk for suicide compared to
acute, episodic depression (e.g. Torpey and Klein, 2008; Arnow
n
Corresponding author.
E-mail address: tforkmann@ukaachen.de (T. Forkmann).
have a chronic course, too. In a community-based cohort study,
Young et al. (2008) found that 51% of all patients with a chronic
depression reported persisting SI within a follow-up interval of 32
months, while only 32% received either appropriate medication or
counseling.
Although it seems plausible to assume that psychological
treatments for depression not only affect depressed affect but also
SI, this assumption has lately been called into question (Cuijpers
et al., 2013). Recent research suggests that suicidal thoughts and
behavior are not merely a part or a symptom of a depressive dis-
order but represent a separate nosological entity (Ahrens and
Linden, 1996; Ahrens et al., 2000; Forkmann et al., 2013; Kendel
et al., 2010; Leboyer et al., 2005; Van Orden et al., 2011). Cuijpers

http://dx.doi.org/10.1016/j.jad.2016.01.047
0165-0327/& 2016 Elsevier B.V. All rights reserved.
52 T. Forkmann et al. / Journal of Affective Disorders 200 (2016) 51–57
et al. (2013) found in a recent meta-analysis only three rando-
mized controlled studies for adult depression in which SI and
suicidal behavior was used as a clearly specified outcome measure.
The pooled results indicated very small and non-significant effects
of psychotherapy for depression on suicidality, but there was not
enough statistical power to consider this a true effect. Moreover, in
a meta-regression analysis with the effect size on SI as dependent
variable and the effect size on depression as predictor, they found
no significant association between both, thus indicating that the
effect of treatments on suicidality was independent from the effect
on depression. In addition to the studies identified by Cuijpers
et al. (2013) two recent studies investigated the effect of psy-
chotherapy for depression on SI. Weitz et al. (2014) found that
Interpersonal Therapy (IPT) had some effectiveness for reducing SI.
However, this effect vanished when controlling for change in de-
pression. A recent pilot study by Ducasse et al. (2014) demon-
strated that Acceptance and Commitment Therapy (ACT), a “third
wave” behavioral therapy including mindfulness elements might
significantly reduce SI. However, its generalizability is limited due
to a small sample size and the absence of a control group.
One further psychotherapeutic approach that may be a treat-
ment option not only for depression but also for SI is Mindfulness-
based Cognitive Therapy (MBCT; Segal et al., 2002) originally de-
signed for relapse prevention of remitted patients with a Major
Depressive Disorder (MDD). MBCT is a group-based program
combining intensive training in meditative practices with cogni-
tive-behavioral elements targeted at depression. MBCT as a treat-
ment option for suicidality is based on the idea, that once a person
has suffered a suicidal crisis, suicidal thoughts are likely to become
reactivated as part of a suicidal mode of mind whenever sad mood
reappears (Williams and Swales, 2004). MBCT may target this
process by enabling people to adopt a different, more decentered
relationship with their own thoughts, feelings and body sensa-
tions, thus preventing these experiences from launching a down-
ward mood spiral that could otherwise lead to another suicidal
crisis. Several studies have shown that MBCT can reduce the risk of
depressive relapse/recurrence in formerly depressed patients (Piet
and Hougaard, 2011). Moreover, Barnhofer et al. (2009) examined
the effects of MBCT plus Treatment as Usual (TAU: continuation of
current medication and appointments with mental health practi-
tioners) in comparison to TAU alone in 28 patients suffering from
chronic depression with a history of SI and suicidal behavior. Re-
sults suggested a significant effect of MBCT on depression but not
on SI. The effect size for the difference between pre- and post-
treatment assessments of SI in the MBCT group was d ¼ 0.48. It
might be that this effect would have reached significance in a
larger sample with sufficient power. Furthermore, in a sample of
patients with residual depressive symptoms, Forkmann et al.
(2014) found a significant reduction in SI in a MBCT group (n ¼64),
whereas no significant changes occurred in a wait-list control-
group (n ¼66) by comparing baseline and post treatment assess-
ments. This interaction effect was independent from the impact of
changes in depression, rumination, and mindfulness skills. How-
ever, change in worry was a significant covariate of the specific
reduction of suicidality in the MBCT group as compared to the
control group. Yet, it remains unclear whether this effect holds
true for a sample of chronic depressed patients.
A specific psychotherapeutic model that has been proposed as
being effective for the treatment of chronic depression is the
Cognitive Behavioral Analysis System of Psychotherapy (CBASP;
McCullough, 2000). Concerning suicidality, the CBASP therapist
received special techniques to deal and cope with suicidal beha-
vior. With the background of the Significant Other History – a list
consisting of significant persons and their impact on the patient's
life – the patient and the therapist could get a deeper under-
standing of the suicidal thoughts and behavior (f. ex.: Because my
father wanted to perform an abortion, I am feeling unwanted and
tired of life). By using the Disciplined Personal Involvement, sui-
cidal patients can learn that their suicidal thoughts and plans
authentically concern and worry the therapist (such as: That you
are thinking about killing yourself, really worries and shocks me).
While using the Interpersonal Discrimination Exercise suicidal
patients can realize that the therapist reacts mostly different in
comparison with the significant others who for example have ig-
nored or punished suicidal behavior. In addition, by conducting
Situational Analysis with situations where suicidal thoughts or
behavior appear patients could learn reconstructing suicidal
thoughts and behavior to overcome helplessness and hopelessness
and to reach their positive desired outcomes (f. ex. instead of
withdrawal and suicidal thoughts or behavior asking for help).
CBASP was originally developed for individual sessions. In recent
years, it has been modified for inpatient settings and group for-
mats (Brakemeier and Normann, 2012; Schramm et al., 2012).
In most of the outpatient studies CBASP proved to be an ef-
fective treatment, especially in combination with medication
(Keller et al., 2000), in patients with childhood trauma (Nemeroff
et al., 2003), early onset (Schramm et al., 2011), and with a long
duration of treatment (Wiersma et al., 2014). In addition, in an
open study with 70 treatment-resistant chronically depressed in-
patients, the CBASP inpatient program significantly reduced de-
pression severity between pre- and posttreatment assessments
with strong effect sizes and high response rates (Brakemeier et al.,
2011, 2015). In a reanalysis of this study, suicidality as measured
with the respective items of the Hamilton Depression Rating Scale
(HAMD) and of the Beck Depression Inventory (BDI) was sig-
nificantly reduced after the inpatient treatment and still after
6 and 12 months (Brakemeier, 2014).
The current study is a secondary analysis of a randomized
controlled trial on the effect of MBCT and CBASP group treatment
on depressive symptomatology in chronic depression conducted at
two treatment sites (trial number NCT01065311; Michalak et al.,
2015). The primary analysis has shown that in the overall sample
as well as at one treatment site, MBCT was no more effective than
TAU in reducing depressive symptoms, while it was significantly
superior to TAU at the other treatment site. CBASP was sig-
nificantly more effective than TAU in reducing depressive symp-
toms in the overall sample and at both treatment sites. However, it
has to be noted that the analysis of differences between sites was a
subsidiary question and it has to be kept in mind that statistical
power for the interpretation of differences between sites is lim-
ited. The direct comparison of the effects of MBCT and CBASP re-
vealed no significant difference but only a statistical trend favoring
CBASP over MBCT. Both treatments had only small to medium
effects on social functioning and quality of life.
The primary aim of the present investigation was to examine
the effects of MBCT and CBASP – both conducted in the format of a
group therapy – on SI compared to a TAU condition in chronic
depressed patients while controlling for changes in other de-
pressive symptoms. On the supposition that suicide behavior lies
on a continuum from ideation through intent and planning to
action, it seems appropriate to choose SI as the dependent variable
of interest.
2. Methods
2.1. Sample
All outpatients had a current major depressive episode (MDE)
as defined by the Diagnostic and Statistical Manual of Mental Dis-
orders (4th ed.; DSM-IV; American Psychiatric Association, 2001)
and experienced depressive symptoms for more than 2 years
 T. Forkmann et al. / Journal of Affective Disorders 200 (2016) 51–57
without remission. We included three subtypes of depressed pa-
tients: (a) patients with chronic major depression (i.e., current
MDE lasting for more than two years); (b) patients meeting criteria
for double depression (current MDE superimposed on an ante-
cedent dysthymic disorder), or (c) patients with current MDE as
part of a recurrent major depression with incomplete recovery
between episodes during the last two years (i.e., depressive
symptoms present during the entire two year period). They were
recruited either by media announcements or through community
health care facilities as well as private practices. Patients were
recruited at two different sites in Germany: the Ruhr megalopolis
and the area of Freiburg im Breisgau. Exclusion criteria that were
used in previous studies on MBCT were applied (e.g., Teasdale
et al., 2000): history of schizophrenia or schizoaffective disorder,
current substance abuse, eating disorder, organic mental disorder,
borderline personality disorder, unable to engage with treatment
for physical, practical, or other reasons. Patients with borderline
personality disorders were excluded because their style of inter-
action might be too difficult to deal with within the group format
of MBCT or CBASP. Mean age of the patients was 54.0 years
(SD ¼13.24, N ¼35, 65.7% female) in the TAU group, 50.2 years
(SD ¼10.5, N ¼35, 63.9% female) in the CBASP group, and 48.4
years (SD ¼11.5, N ¼36, 58.2% female) in the MBCT group. Further
characteristics of the sample and a CONSORT flow diagram illus-
trating patient flow are presented in detail elsewhere (Michalak
et al., 2015).
2.2. Sampling procedures
All study procedures were approved by the Research Ethics
Committee of the German Psychological Association (JM 72009)
and all participants signed an informed consent form. All in-
dividuals interested in participating in the study took part in a
telephone screening based on the mood disorder module of the
German version of the Structured Clinical Interview for DSM-IV
(SCID; Wittchen et al., 1997). Patients who seemed eligible were
invited for an extended diagnostic interview. Psychiatric diagnoses
were assessed using SCID Axis I and Axis II disorders. All diagnostic
evaluations were conducted by trained and certified clinical psy-
chologists and were reviewed by senior study investigators (J.M. &
E.S.).
2.3. Randomization
At each trial site, patients were randomly assigned to either
TAU or to receive – in addition to TAU – either MBCT or CBASP.
Block randomization (block size ¼ 6) to the three conditions was
performed by an independent allocator using a computer-gener-
ated list of random numbers. After patient eligibility was assessed
and informed consent was obtained, patients were formally en-
rolled in the study. Thirty-six patients were allocated to MBCT, 35
to CBASP, and 35 to TAU.
2.4. Treatments
2.4.1. Treatment as Usual (TAU)
All patients were instructed to be in individual care of either a
psychiatrist or a licensed psychotherapist (not member of the
study team) during the study period. If patients were already in
psychiatric/psychotherapeutic individual treatment at study in-
take, they continued their treatment with this psychiatrist/psy-
chotherapist. Patients were encouraged to continue any current
medication and to attend appointments with their psychiatrist/
psychotherapist. There were no restrictions on other forms of
supplementary treatment. Detailed information on the balance of
active ingredients of TAU (medication, individual psychotherapy,
53
appointments with psychiatrists) in all three groups is given in
Michalak et al. (2015). In general, no differences between the
groups were found.
2.4.2. Mindfulness-based Cognitive Therapy (MBCT)
The treatment protocol followed the MBCT manual developed
by Segal et al. (2002). Some minor alterations were made to adapt
the program to chronically depressed patients. More specifically,
the presence of acute symptoms was addressed. The program
consists of an individual pre-class interview and eight weekly 2.5-
h group sessions. Group size was restricted to six patients per class
in the present trial to parallel group size in both psychotherapy
conditions.
2.4.3. Cognitive Behavioral Analysis System for Psychotherapy
(CBASP)
The CBASP treatment protocol followed the manual developed
by McCullough (2000) modified for the group setting by Schramm
et al. (2012). The CBASP program consisted of two individual
treatment sessions and eight weekly 2.5-h group sessions. The
main modifications of the individual format included the deriva-
tion of the transference hypothesis with regard to the group and to
use Disciplined Personal Involvement more sparsely or replace it
by using the Kiesler Circle model (Kiesler, 1983) in an educative
and structuring way. In order to cover Situational Analyses of all
group members, the number of participants was limited to six.
2.5. Measures
2.5.1. Depressive symptoms
Depressive symptoms were assessed using two instruments:
the HAMD and the BDI. The 24-item HAMD (Hamilton, 1960, 1967)
is a widely used interview-based measure of severity of depressive
symptoms covering a range of affective, behavioral, and biological
symptoms. It was the primary outcome measure in the main study
reported on in Michalak et al. (2015). The HAMD was administered
at baseline and post-treatment after the eight weeks treatment
phase by five trained doctoral-level psychologists. To enhance re-
liability of the assessment the Structured Interview Guide for the
HAMD (Moberg et al., 2001) was applied. A sample of 36 inter-
views from the baseline assessment were second-rated for the
HAMD by an independent rater to yield an interrater correlation of
r (34) ¼.97, p o 0.001. To maintain blindness of the raters, patients
were instructed at the beginning of each rating session not to
mention their treatment condition or their psychotherapist during
the interviews.
The BDI (Beck et al., 1996; Hautzinger et al., 2006) was applied
to assess depressive symptoms by self-report. The BDI is a widely
used 21-item measure covering affective, cognitive, and biological
symptoms of depression with good psychometric properties. In
the current sample, the BDI showed good reliability (Cronbach's
α ¼.81).
The suicide items were excluded from the HAMD and the BDI
sum scores to avoid overlapping and thus part-whole correlations.
2.5.2. Suicidal ideation (SI)
SI was measured by using (1.) the respective item from the
interviewer based HAMD (Item 3; Hamilton, 1960) as well as (2.)
the respective item from the self-rating form of the German ver-
sion of the BDI (Item 9; Beck et al., 1996; Hautzinger et al., 2006).
The suicide item of the HAMD and the suicide item of the BDI have
both repeatedly been used to measure suicidality in clinical stu-
dies (e.g., Larsson et al., 1991; Lewinsohn et al., 1993; Weitz et al.,
2014). In terms of convergent validity, the BDI and HAMD suicid-
ality items have been shown to be highly correlated (r¼ .80) with
the first five items of the Beck Scale for Suicide Ideation (Ballard
54 T. Forkmann et al. / Journal of Affective Disorders 200 (2016) 51–57

et al., 2015; see also Beck et al., 1997). For the HAMD suicide item,
convergent validity has also been shown with the Adult Suicide
Ideation Questionnaire (Reynolds, 1991). Predictive validity of both
the BDI and the HAMD suicide item for later suicidal behavior has
been shown to be good (Brown et al., 2000).
2.6. Data analyses
Analyses were conducted on the intention to treat sample.
Missing data was accounted for by using the last observation
carried forward (LOCF). Analyses of variance (ANOVA) and χ2 tests
were used to examine whether treatment groups differed on sui-
cidality at baseline. We conducted paired-samples t-tests and
calculated Cohen's effect size d to analyse whether suicidality as
measured with the HAMD and BDI suicide items were lower after
treatment than at baseline within each single treatment group
(Cohen, 1988). In order to reduce sampling error, effect sizes have
been corrected using a factor provided by Hedges and Olkin
(1985). Multivariate linear regression analyses were used to test
for significant differences in reduction of suicidal symptoms be-
tween the treatment groups (MBCT, CBASP) and TAU. Change in
suicidal symptoms measured with the HAMD and the BDI suicide
items were the dependent variables. For the predictor variables,
dummy variables were used for the MBCT and the CBASP groups
with TAU as reference condition. Analyses were controlled for
changes in depression (measured with the HAMD and BDI total
scores exclusive of the respective suicide items) (model 1). In a
secondary set of multivariate linear regression analyses, we com-
pared the treatment conditions MBCT and CBASP directly to each
other. MBCT was the reference condition.
All analyses were performed using IBM SPSS statistics 20.
3. Results
3.1. Primary outcome analyses
According to the HAMD suicide item, 61.1% of the participants
showed any kind of SI in the MBCT group, 65.7% in the CBASP
group and 62.9% in the TAU group (χ2 ¼.16, p¼ .92). According to
the BDI suicide item, 77.8% of the participants in the MBCT group
reported any kind of SI, 47.1% in the CBASP group and 75.8% in the
TAU group (χ2 ¼9.15, p ¼.01). In the MBCT group, 5 participants
reported one or more past suicide attempt, in the CBASP group
8 and in the TAU group 6 (χ2 ¼.75, p ¼.69).
Means and standard deviations of the HAMD and the BDI sui-
cide items and the HAMD and BDI total scores without the in-
clusion of the suicide items pre and post treatment are displayed
in Table 1. Baseline scores on the HAMD suicide item did not differ
between treatment groups (F(2103) ¼ 0.48; p ¼0.62). However, as
can also be seen in Fig. 1, significant differences between treat-
ment groups at baseline were found for the BDI suicide item (F
(2100) ¼4.30; p ¼0.02).
Paired-samples t-tests revealed that both MBCT and CBASP
reduced suicidal symptoms if measured with the HAMD suicide
item. Suicidality as measured with the BDI suicide item was sig-
nificantly reduced by MBCT but not by CBASP. Effect sizes (Cohen's
d) showed medium effects of MBCT on SI both measured with the
HAMD suicide item (d ¼0.30) and the BDI suicide item (d ¼0.34).
Moreover, MBCT reduced depression both measured with the
HAMD (d ¼0.38) and BDI (d ¼0.52). CBASP affected SI only when
measured with the HAMD suicide item (d ¼0.50) but not when
measured with the BDI suicide item (d ¼0.07). However, due to
significantly lower scores on the BDI suicide item in the CBASP
group than in the MBCT group at baseline, this could be due to a
floor effect in the CBASP group. Again, effects on depression were
of medium size (HAMD: d ¼0.78; BDI: d ¼0.60). TAU showed vir-
tually no effects on suicidality but on depression when measured
with the HAMD (d¼ 0.35; Table 1).
3.2. Incremental effects of treatment conditions compared to TAU;
dependent variable: HAMD suicide item
Dummy-coded multivariate linear regression analyses were
conducted with dummy variables for the MBCT and the CBASP
groups with TAU as reference condition. Significant effects of these
dummy variables on the dependent variable indicated that the
respective treatment condition had a significantly larger effect on
SI than TAU. Differential effects of treatment conditions were
found if SI was measured with the HAMD suicide item (F(3102) ¼
12.14; p o0.001; Table 2). Both MBCT (model 1: β ¼ 0.24; p ¼0.02)
and CBASP (β ¼ 0.27; p ¼0.01) reduced SI significantly as compared
to TAU while controlling for changes in depression (model 1).
When comparing MBCT and CBASP directly to each other with
MBCT as reference condition, no significantly different effect of
CBASP (β ¼.07; p ¼.56) compared to MBCT was found (model 2: F
(268) ¼5.64, p o0.01) (Fig. 2).
3.3. Incremental effects of treatment conditions compared to TAU;
dependent variable: BDI suicide item
There was no differential effect of the treatment conditions as
compared to TAU on suicidal symptoms when assessed with the
BDI suicide item while controlling for changes in depression
(model 1) (Table 2). Neither MBCT (model 1: β ¼ 0.04; p¼ 0.72) nor
CBASP (model 1: β ¼  0.10; p¼ 0.32) showed an incremental effect
on the reduction of SI as assessed with the BDI suicide item
compared to TAU. However, when compared to each other with
MBCT as reference condition, CBASP showed a significantly smaller
effect on suicidality (β ¼  .25; p¼ .03) compared to MBCT (model

Table 1
Descriptive statistics and pre-post changes of suicidal ideation in the MBCT, CBASP and TAU group.

MBCT CBASP TAU

Pre Posta Pre Posta Pre Posta

M SD M SD t p d M SD M SD t p d M SD M SD t p d

BDI suicide item 0.86 0.59 0.64 0.68 2.26 0.03 0.34 0.50 0.56 0.46 0.66 0.70 0.49 0.07 0.91 0.72 0.88 0.77 0.49 0.62 0.04
HAMD suicide item 0.83 0.81 0.58 0.84 2.17 0.04 0.30 1.03 0.89 0.57 0.92 4.12 0.00 0.50 0.89 0.90 1.06 1.00  1.23 0.23  0.18
b
HAMD total 23.03 6.27 19.69 10.27 2.63 0.01 0.38 24.71 6.69 18.03 9.76 5.50 0.00 0.78 23.11 6.33 20.91 7.86 2.87 0.01 0.35
BDI totalb 31.31 9.55 25.40 12.70 3.26 0.00 0.52 30.27 7.52 24.43 11.34 3.55 0.00 0.60 29.77 7.42 28.64 10.24 1.65 0.11 0.12

a
Missing data was accounted for by using the last observation carried forward (LOCF). MBCT: Mindfulness-based Cognitive Therapy; CBASP: Cognitive Behavioral
Analysis System of Psychotherapy; TAU: Treatment as usual; BDI: Beck Depression Inventory; HAMD: Hamilton Depression Rating Scale; M: mean; SD: standard deviation.
b
suicide item excluded.
 T. Forkmann et al. / Journal of Affective Disorders 200 (2016) 51–57 55

Fig. 1. Change in suicidal ideation as measured with the BDI suicide item in the Fig. 2. Change in suicidal ideation as measured with the HAMD suicide item in the
different treatment conditions. different treatment conditions.

2: F(267)¼7.08, p o0.01).
4. Discussion
The results of the present study showed that both MBCT added
to TAU and CBASP added to TAU have an additional effect on sui-
cidality when measured with the HAMD suicide item. This effect
was robust when controlling for changes in other depression
symptoms. Sizes of the effects of MBCT and CBASP on suicidality
were similar to those reported in prior studies (Barnhofer et al.,
2009; Forkmann et al., 2014). However, it has to be noted that
despite being robust, these effects of MBCT and CBASP on suicid-
ality were not large and it remains an open question whether both
treatments have similar effects on suicidal behavior. Furthermore,
future research must show whether MBCT and CBASP aids in
preventing suicidal relapse.
A different picture emerged when suicidality was assessed via
self-report (BDI suicidality item). In the MBCT group, a significant
reduction in suicidality was found. However, this effect vanished
by controlling for changes in depression. One might speculate
whether this could potentially be caused by multiple testing or
whether it could be interpreted in line with the common as-
sumption that suicidality is just an epiphenomenon of depression.
In the CBASP group, no significant change in suicidality was found
when assessed with the BDI suicidality item. This might be at least
partly due to a floor effect. In the CBASP group, suicidality was
significantly lower than in the MBCT group at baseline which
might have precluded significant effects in the CBASP group. When
comparing both treatments directly to each other no different
effect on suicidality was found when suicidality was measured
with the HAMD suicide item. However, when the BDI suicide item
was chosen as dependent variable, MBCT showed a significantly
larger effect on suicidality than CBASP.
The divergence between results of HAMD and BDI is note-
worthy. Several reasons are conceivable. Generally, prior research
has repeatedly shown that the HAMD appears to be more sensitive
to change than the BDI (see e.g., Schneibel et al., 2012; Sayer et al.,
1993). Moreover, Schneibel et al. (2012) found that low extraver-
sion and high neuroticism were associated with higher endorse-
ment of depressive symptoms on the BDI as compared to the
HAMD. In our own data, the discrepancy between HAMD and BDI
might be associated with a potential floor effect. For both HAMD
and BDI suicide items the mean ratings at baseline were o1.
However, as Table 1 shows, HAMD data had slightly larger variance
than BDI data. This larger variance could have facilitated finding a
slightly larger effect of treatments. Second, verbal anchors of the
steps of the rating scales differ. For the BDI item, step “1” refers to
SI without concrete intent or plans to commit suicide. However,
for the HAMD item step “1” refers to more general feelings that life
is not worth living. Step “2” of the BDI item refers to the wish to
commit suicide, whereas step “2” of the HAMD item refers to the
wish to be dead without the concrete implication of suicide. Thus,
steps “1” and “2” of the HAMD item might have been slightly easier,
i.e., probability to endorse this answer option might have been
higher than for the corresponding answer option of the BDI item.
Indeed, at baseline, for a total of n¼20 participants option two of
the HAMD item was endorsed and only n ¼3 participants endorsed
option two of the BDI item. A third explanation for the divergent
results is that effects of social desirability could have been more

Table 2
Effects of MBCT and CBASP on change in HAMD and BDI suicide scores compared to TAU (model 1) and compared to each other (model 2).

Dependent variable: change in HAMD suicide score Dependent variable: change in BDI suicide score

B SD β t p adj. R2 B SD β t p adj. R2

Model 1 Intercept  0.33 0.12  2.77 0.01 0.24  0.02 0.10  0.18 0.86 0.26
change in depression 0.05 0.01 0.40 4.50 0.00 0.04 0.01 0.52 5.91 0.00
MBCT vs. TAU 0.39 0.16 0.24 2.46 0.02 0.05 0.13 0.04 0.36 0.72
CBASP vs. TAU 0.44 0.17 0.27 2.66 0.01  0.13 0.13  0.10  1.00 0.32

Model 2 Intercept 0.11 0.12 0.97 0.33 0.12 0.10 0.09 1.08 0.28 0.15
change in depression 0.04 0.01 0.36 3.07 0.00 0.03 0.01 0.40 3.51 0.00
CBASP vs. MBCT 0.09 0.16 0.07 0.58 0.56  0.27 0.12  0.25  2.19 0.03

Note: MBCT: Mindfulness-based Cognitive Therapy; CBASP: Cognitive Behavioral Analysis System of Psychotherapy; TAU: Treatment as usual; HAMD: Hamilton Depression
Rating Scale; BDI: Beck Depression Inventory.
56 T. Forkmann et al. / Journal of Affective Disorders 200 (2016) 51–57
powerful in the clinician rated HAMD than in the self-rated BDI.
This effect could have led to an artificial reduction of reported
symptom burden at post assessment in the HAMD, because par-
ticipants could have suspected that the interviewer expected them
to have benefited from treatment. However, this supposition is
undermined by the fact that for the total HAMD treatment effects
were not larger than for the total BDI (see Table 1). In sum, the
divergence between results in clinician rated and self-reported
suicidality should be further investigated. Generally, the applica-
tion of both self-report and clinician rated instruments should be
recommended.
A common mechanism of action that could have accounted for
the incremental effects on suicidality of both MBCT and CBASP in
addition to TAU compared to TAU alone is peer support. MBCT and
CBASP consisted of eight weekly group sessions and one or two
additional individual sessions. TAU contained no formal group-
based intervention but only regular psychiatric / psychother-
apeutic treatments. Social isolation is one of the strongest risk
factors for SI and in contrast, marriage, children, and a greater
number of friends are associated with decreased suicide risk (Van
Orden et al., 2010). Additionally, according to contemporary the-
ories on the development of SI and suicidal behavior – such as the
Interpersonal Theory of Suicide (Joiner, 2005), or the Integrated
Motivational-volitional Model of Suicidal Behavior (O'Connor,
2011) – the absence of reciprocal care is one key component of a
sense of thwarted belongingness that may act as a motivational
moderator or even direct predictor of increases in SI. Thus, the
treatments delivered in groups – such as in this case MBCT and
CBASP – may have reduced SI by increasing the feeling of social
belongingness of the participants. However, results of a meta-
analysis by Tarrier et al. (2008) suggest that for cognitive-beha-
vioral interventions for suicidal thoughts and behavior generally
even larger effects for individual than for group-based treatments
could be found. Thus, the effects of MBCT and CBASP on suicidality
as reported in the present study are probably not reducible to a
simple effect of the group-setting that would have been found in
any other group-based treatment, too.
In all models investigated, change in depression was a sig-
nificant predictor of change in SI. This is in line with the as-
sumption that depression and suicidality are closely related psy-
chopathological syndromes. However, our results also show that
both MBCT and CBASP can have an additional effect on suicidality
above that of change in depression. This result corroborates the
view that it cannot be assumed that depression focused inter-
ventions will be effective in reducing suicidal thoughts and be-
havior (Cuijpers et al., 2013) and that suicidal thoughts and be-
havior are not merely a part or a symptom of a depressive disorder
but represent a separate nosological entity (Ahrens and Linden,
1996; Ahrens et al., 2000; Forkmann et al., 2013; Kendel et al.,
2010; Leboyer et al., 2005; Van Orden et al., 2011).
4.1. Strengths and limitations
Some strengths and limitations should be kept in mind when
appreciating the results of the present study. Treatments were
delivered mostly group-based and with limited duration which
might have had an impact on outcome. Patients were encouraged
to continue any current medication and to attend appointments
with their psychiatrist/psychotherapist. There were no restrictions
on other forms of supplementary treatment. Suicidality was
measured with two single items. Although this is common practice
in this area of research (Ladwig et al., 2010; Weitz et al., 2014;
Krajniak et al., 2013), use of psychometric scales for SI would be
desirable and may improve reliability of assessments in future
investigations. However, it has to be noted that in contrast to prior
studies (e.g., Forkmann et al., 2014; Ladwig et al., 2010) we
included two single items, one self- and one clinician-rated item.
Power of analyses was reasonable for the comparison of the group
therapy conditions (MBCT þ TAU and CBASP þ TAU) with TAU
alone, but might have been insufficient for the direct comparison
of CBASP to MBCT (secondary analysis). Although some effort has
been invested to maintain blindness of the raters (i.e., patients
were instructed not to mention their treatment condition or their
psychotherapist during the interviews, ratings and therapy were
conducted in different buildings) there was no direct measure of
blindness. Strengths of the study are the randomized-controlled
design, the inclusion of a control group, and two active treatments.
4.2. Summary and conclusion
In conclusion, results of the present randomized-controlled
trial suggest that both MBCT and CBASP may be successful psy-
chotherapeutic approaches for the group-based treatment of SI in
chronically depressed patients. Incremental effects of both treat-
ments above the effect of TAU were dependent on the way sui-
cidality was assessed (self- vs. clinician-rated) but were in-
dependent from changes in depression. Future research should try
to reveal variables that mediate or moderate the effects of MBCT
and CBASP on SI in order to facilitate a deeper understanding of
the clinical mechanisms of action. Because all effects were of small
to medium size and were independent of effects from other de-
pression symptoms, the present results warrant the application of
such psychotherapeutical treatment strategies like MBCT and
CBASP for SI in patients with chronic depression.
Trial registration and funding
This trial was registered (NCT01065311) and was funded by the
German Science Foundation (DFG: Mi 700/4-1).
References
Ahrens, B., Linden, M., 1996. Is there a suicidality syndrome independent of specific
major psychiatric disorder? Results of a split half multiple regression analysis.
Acta Psychiatr. Scand. 94, 79–86.
Ahrens, B., Linden, M., Zaske, H., Berzewski, H., 2000. Suicidal behavior - symptom
or disorder? Compr. Psychiatr. 41, 116–121.
American Psychiatric Association, 2001. Diagnostic & Statistical Manual of Mental
Disorders. American Psychiatric Association, Washington, DC.
Arnow, B.A., Manber, R., Blasey, C., Klein, D.N., Blalock, J.A., Markowitz, J.C., Roth-
baum, B.O., Rush, A.J., Thase, M.E., Riso, L.P., Vivian, D., McCullough Jr., J.P.,
Keller, M.B., 2003. Therapeutic reactance as a predictor of outcome in the
treatment of chronic depression. J. Consult. Clin. Psychol. 71, 1025–1035.
Arsenault-Lapierre, G., Kim, C., Turecki, G., 2004. Psychiatric diagnoses in 3275
suicides: a meta-analysis. BMC Psychiatry 4, 37.
Ballard, E.D., Luckenbaugh, D.A., Richards, E.M., Walls, T.L., Brutsche, N.E., Ameli, R.,
et al., 2015. Assessing measures of suicidal ideation in clinical trials with a
rapid-acting antidepressant. J. Psychiatr. Res. 68, 68–73.
Barnhofer, T., Crane, C., Hargus, E., Amarasinghe, M., Winder, R., Williams, J., 2009.
Mindfulness-based cognitive therapy as a treatment for chronic depression: a
preliminary study. Behav. Res. Ther. 47, 366–373.
Beck, A.T., Steer, R.A., Brown, G.K., 1996. Manual for the Beck Depression Inventory,
2nd ed. The Psychological Corporation, San Antonio.
Beck, A.T., Brown, G.K., Steer, R.A., 1997. Psychometric characteristics of the Scale
for Suicide Ideation with psychiatric outpatients. Behav. Res. Ther. 35,
1039–1046.
Borges, G., Nock, M., Abad, J., Hwang, I., Sampson, N., Alonso, J., 2010. Twelve-month
prevalence of and risk factors for suicide attempts in the World Health Orga-
nization World Mental Health Surveys. J. Clin. Psychiatry 71, 1617–1628.
Brakemeier, E.L., Engel, V., Schramm, E., Zobel, I., Schmidt, T., Hautzinger, M., Berger,
M., Normann, C., 2011. Feasibility and outcome of cognitive behavioral analysis
system of psychotherapy (CBASP) for chronically depressed inpatients: a pilot
study. Psychother. Psychosom. 80, 191–194.
Brakemeier, E.-L., Normann, C., 2012. Praxisbuch CBASP. Behandlung chronischer
Depression. Beltz, Weinheim.
Brakemeier, E.-L., Radtke, M., Engel, V., Zimmermann, J., Tuschen-Caffier, B.,
Schramm, E., Hautzinger, M., Berger, M., Normann, C., 2015. Overcoming
treatment-resistance in chronic depression: outcome and feasibility of the
 T. Forkmann et al. / Journal of Affective Disorders 200 (2016) 51–57
Cognitive Behavioral Analysis System of Psychotherapy as an inpatient treat-
ment program. Psychother. Psychosom. 84, 51–56.
Brakemeier, E.-L., 2014. Wie beeinflusst eine intensive, stationäre CBASP-Therapie
die Suizidalität von therapieresistenten, chronisch depressiven Patienten kurz-
und langfristig? Vortrag im Symposium: Neue psychotherapeutische Be-
handlungsstrategien der Depression: Vorhersage und Behandlung von Suizi-
dalität (Vorsitz: T. Forkmann) auf dem 32. Symposium Klinische Psychologie
und Psychotherapie der DGPS Fachgruppe Klinische Psychologie und Psy-
chotherapie. Braunschweig, 29.-31.05.2014.
Brown, G.K., Beck, A.T., Steer, R.A., Grisham, J.R., 2000. Risk factors for suicide in
psychiatric outpatients: a 20-year prospective study. J. Consult. Clin. Psychol.
68, 371–377.
Cohen, J., 1988. Statistical Power for the Behavioural Science, 2nd ed. Erlbaum,
Hillsdale, New Jersey.
Cuijpers, P., de Beurs, D.P., van Spijker, B.A., Berking, M., Andersson, G., Kerkhof, A.J.,
2013. The effects of psychotherapy for adult depression on suicidality and
hopelessness: a systematic review and meta-analysis. J. Affect. Disord. 144,
183–190.
Ducasse, D., René, E., Béziat, S., Gullaume, S., Courtet, P., Olié, E., 2014. Acceptance
and Commitment Therapy for management of suicidal patients: a pilot study.
Psychother. Psychosom. 83, 374–376.
Forkmann, T., Gauggel, S., Spangenberg, L., Brahler, E., Glaesmer, H., 2013. Dimen-
sional assessment of depressive severity in the elderly general population:
psychometric evaluation of the PHQ-9 using Rasch Analysis. J. Affect. Disord.
148, 323–330.
Forkmann, T., Wichers, M., Geschwind, N., Peeters, F., van Os, J., Mainz, V., Collip, D.,
2014. Effects of mindfulness-based cognitive therapy on self-reported suicidal
ideation: results from a randomised controlled trial in patients with residual
depressive symptoms. Compr. Psychiatry 55, 1883–1890.
Hamilton, M., 1960. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry
23, 56–62.
Hamilton, M., 1967. Development of a rating scale for primary depressive illness. Br.
J. Soc. Clin. Psychol. 6, 278–296.
Hautzinger, M., Keller, F., Kühner, C., 2006. Beck Depressions-Inventar Revision:
Manual. Harcourt, Frankfurt am Main.
Hedges, L.V., Olkin, I., 1985. Statistical Methods for Meta-Analysis. Academic Press,
Orlando.
Joiner, T., 2005. Why People Die by Suicide. Harvard University Press, Cambridge,
Massachusetts.
Keller, M.B., McCullough, J.P., Klein, D.N., Arnow, B., Dunner, D.L., Gelenberg, A.J.,
Markowitz, J.C., Nemeroff, C.B., Russell, J.M., Thase, M.E., Trivedi, M.H., Zajecka,
J., 2000. A comparison of nefazodone, the cognitive behavioral-analysis system
of psychotherapy, and their combination for the treatment of chronic depres-
sion. N. Engl. J. Med. 342, 1462–1470.
Kendel, F., Wirtz, M., Dunkel, A., Lehmkuhl, E., Hetzer, R., Regitz-Zagrosek, V., 2010.
Screening for depression: rasch analysis of the dimensional structure of the
PHQ-9 and the HADS-D. J. Affect. Disord. 112, 241–246.
Kiesler, D.J., 1983. The 1982 interpersonal circle – a taxonomy for complementarity
in human transactions. Psychol. Rev. 90, 185–214.
Krajniak, M., Miranda, R., Wheeler, A., 2013. Rumination and pessimistic certainty
as mediators of the relation between lifetime suicide attempt history and fu-
ture suicidal ideation. Arch. Suicide Res. 17, 196–211.
Ladwig, K.H., Klupsch, D., Meisinger, C., Baumert, J., Erazo, N., Schneider, A.,
Wichmann, H.-E., 2010. Gender differences in risk assessment of death wishes
and suicidal ideation in the community: results from the KORA Augsburg F3
study with 3079 men and women, 35 to 84 years of age. J. Nerv. Ment. Dis. 198,
52–58.
Larsson, B., Melin, L., Breitholtz, E., Andersson, G., 1991. Short-term stability of
depressive symptoms and suicide attempts in Swedish adolescents. Acta Psy-
chiatr. Scand. 83, 385–390.
Leboyer, M., Slama, F., Siever, L., Bellivier, F., 2005. Suicidal disorders: a nosological
entity per se? Am. J. Med. Genet. C 133, 3–7.
Lewinsohn, P.M., Rohde, P., Seeley, J.R., 1993. Psychosocial characteristics of ado-
lescents with a history of suicide attempt. J. Am. Acad. Child Psychiatry 32,
60–68.
McCullough, J.P., 2000. Treatment of Chronic Depression. Cognitive Behavioral
Analysis System of Psychotherapy. Guilford Press, New York.
Michalak, J., Schutze, M., Heidenreich, T., Schramm, E., 2015. A randomized
57
controlled trial on the efficacy of Mindfulness-Based Cognitive Therapy and a
group version of Cognitive Behavioral Analysis System of Psychotherapy for
chronically Depressed Patients. J. Consult. Clin. Psychol. 83, 951–963.
Moberg, P.J., Lazarus, L.W., Mesholam, R.I., Bilker, W., Chuy, I.L., Neyman, I., Mark-
vart, V., 2001. Comparison of the standard and structured interview guide for
the Hamilton Depression Rating Scale in depressed geriatric inpatients. Am. J.
Geriat. Psychiatry 9, 35–40.
Nemeroff, C.B., Heim, C.M., Thase, M.E., Klein, D.N., Rush, A.J., Schatzberg, A.F., Ni-
nan, P.T., McCullough Jr., J.P., Weiss, P.M., Dunner, D.L., Rothbaum, B.O., Korn-
stein, S., Keitner, G., Keller, M.B., 2003. Differential responses to psychotherapy
versus pharmacotherapy in patients with chronic forms of major depression
and childhood trauma. Proc. Natl. Acad. Sci. 100, 14293–14296.
O'Connor, R.C., 2011. Towards an integrated motivational-volitional model of sui-
cide behavior. In: O’Connor, R.C., Platt, S., Gordon, J. (Eds.), International
Handbook of Suicide Prevention: Research Policy and Practice. John Wiley &
Sons, Ltd., Oxford, pp. 181–198.
Piet, J., Hougaard, E., 2011. The effect of mindfulness-based cognitive therapy for
prevention of relapse in recurrent major depressive disorder: a systematic re-
view and meta-analysis. Clin. Psychol. Rev. 31, 1032–1040.
Reynolds, W.M., 1991. Adult Suicidal Ideation Questionnaire (ASIQ) – Professional
Manual. Psychological Assessment Resources, Odessa.
Sayer, N.A., Sackeim, H.A., Moeller, J.R., Prudic, J., Devanand, D.P., Coleman, E.A.,
Kiersky, J.E., 1993. The relations between observer-rating and self-report of
depressive symptomatology. Psychol. Assess. 5, 350–360.
Schneibel, R., Brakemeier, E.-L., Wilbertz, G., Dykierek, P., Zobel, I., Schramm, E.,
2012. Sensitivity to detect change and the correlation of clinical factors with the
Hamilton Depression Rating Scale and the Beck Depression Inventory in de-
pressed inpatients. Psychiatry Res. 198, 62–67.
Schramm, E., Fangmeier, R., Brakemeier, E.-L., 2012. CBASP in der Gruppe. Schat-
tauer, Stuttgart.
Schramm, E., Zobel, I., Dykierek, P., Kech, S., Brakemeier, E.L., Kulz, A., Berger, M.,
2011. Cognitive behavioral analysis system of psychotherapy versus inter-
personal psychotherapy for early-onset chronic depression: a randomized pilot
study. J. Affect. Disord. 129, 109–116.
Segal, Z.V., Williams, J.M.G., Teasdale, J.D., 2002. Mindfulness-Based Cognitive
Therapy for Depression: A New Approach to preventing Relapse. Guilford, New
York.
Tarrier, N., Taylor, K., Gooding, P., 2008. Cognitive-behavioral interventions to re-
duce suicide behavior: a systematic review and meta-analysis. Behav. Modif. 32,
77–108.
Teasdale, J.D., Segal, Z.V., Williams, J.M.G., Ridgeway, V.A., Soulsby, J.M., Lau, M.A.,
2000. Prevention of relapse/recurrence in major depression by mindfulness-
based cognitive therapy. J. Consult. Clin. Psychol. 68, 615–623. http://dx.doi.org/
10.1037/0022–006X.68.4.615.
Torpey, D.C., Klein, D.N., 2008. Chronic depression: update on classification and
treatment. Curr. Psychiatry Rep. 10, 458–464.
Van Orden, K.A., Witte, T.K., Holm-Denoma, J., Gordon, K.H., Joiner Jr., T.E., 2011.
Suicidal behavior on Axis VI: clinical data supporting a sixth Axis for DSM-V.
Crisis 32, 110–113.
Van Orden, K.A., Witte, T.K., Cokrowicz, K.C., Braithwaite, S.R., Selby, E.A., Joiner, T.E.,
2010. The interpersonal theory of suicide. Psychol. Rev. 117, 575–600.
Weitz, E., Hollon, S.D., Kerkhof, A., Cuijpers, P., 2014. Do depression treatments
reduce suicidal ideation? The effects of CBT, IPT, pharmacotherapy, and placebo
on suicidality. J. Affect. Disord. 167, 98–103.
WHO, 2014. Preventing Suicide. A Global Imperative. WHO Press, Genf.
Wiersma, J.E., Van Schaik, D.J., Hoogendorn, A.W., Dekker, J.J., Van, H.L., Schoevers,
R.A., Blom, M.B., Maas, K., Smit, J.H., McCullough Jr., J.P., Beekamn, A.T., Van
Oppen, P., 2014. The effectiveness of the cognitive behavioral analysis system of
psychotherapy for chronic depression: a randomized controlled trial. Psy-
chother. Psychosom. 83, 263–269.
Williams, J.M., Swales, M., 2004. The use of mindfulness-based approaches for
suicidal patients. Arch. Suicide Res. 8, 315–329.
Wittchen, H.-U., Zaudig, M., Fydrich, T., 1997. Strukturiertes Klinisches Interview für
DSM-IV. Hogrefe, Göttingen.
Young, A.S., Klap, R., Shoai, R., Wells, K., 2008. Persistent depression and anxiety in
the United States: prevalence and quality of care. Psychiatr. Serv. 59,
1391–1398.