KleinHofmeijer PDF

m ple x an x ie ty
Co
disorders
Risk factors, underlying mechanisms
and treatment enhancement
Mieke Klein Hofmeijer-Sevink

The research reported in this thesis was financially supported with a grant from
ZonMw, the Netherlands Organization for Health Research and development
(grant number 171001007), and Fonds NutsOhra.
Cover design: Ruben Boerhof (www.yellowlemmontree.nl)

Lay out: Juliet Campfens (persoonlijkproefschrift.nl)
Printed by Ipskamp drukkers, Enschede, the Netherlands
ISBN: 978-94-6259-900-0
No part of this book may be reproduced in any form without permission of the author.
© 2015, Mieke Klein Hofmeijer-Sevink, Amersfoort, the Netherlands.
Complex anxiety disorders:
Risk factors, underlying mechanisms
and treatment enhancement
Complexe angststoornissen:
risico factoren, onderliggende mechanismen
en nieuwe behandelopties
(met een samenvatting in het Nederlands)
PROEFSCHRIFT
ter verkrijging van de graad van doctor
aan de Universiteit Utrecht,
op gezag van de rector magnificus
prof. dr. G.J. van der Zwaan,
ingevolge het besluit
van het college voor promoties
in het openbaar te verdedigen op
vrijdag 5 februari 2016 te 10.30 uur
door
Mieke Klein Hofmeijer-Sevink
geboren op 7 mei 1980 te Gorssel

Promotoren:
Prof. dr. M.A. van den Hout
Prof. dr. A.J.L.M. van Balkom
Copromotoren:
Dr. D.C. Cath
Dr. H.J.G.M. van Megen
Content
Chapter 1 General introduction 6
Chapter 2.1 Clinical relevance of comorbidity in anxiety 24

disorders: A report from the Netherlands Study of
Depression and Anxiety (NESDA).
Chapter 2.2 Clinical relevance of comorbidity in Obsessive 40

Compulsive Disorder: The Netherlands OCD
Association Study.
Chapter 2.3 Occurrence and predictive value of clinically 60

relevant obsessive-compulsive symptoms in anxiety
and depressive disorders.
Chapter 3 Expectancy bias and treatment outcome in patients 76

with panic disorder and agoraphobia.
Chapter 4 D-cycloserine does not enhance exposure therapy 90

in patients with panic disorder with agoraphobia: a
double-blind, randomized controlled trial.
Chapter 5 Summary and general discussion 108
Chapter 6 Nederlandse samenvatting (Dutch summary) 124
Dankwoord 132
Curriculum vitae, publicaties, voordrachten 136

Chapter 1
Introduction
Chapter 1
INTRODUCTION
Anxiety disorders are common psychiatric disorders with an estimated lifetime preva-
lence of 19.3 % in the Netherlands1 and, according to DSM-IV criteria, include panic dis-
order (with or without agoraphobia), social phobia, specific phobia, agoraphobia (without
a history of panic disorder), generalized anxiety disorder (GAD), obsessive-compulsive
disorder (OCD), posttraumatic stress disorder (PTSD), and acute stress disorder.2
Complex anxiety disorders are disorders with high severity, comorbidity, or treat-
ment resistance. Since complex anxiety disorders tend to develop a chronic course
and the prevalence rate of chronicity in anxiety is high, research on putative risk fac-
tors, potential underlying mechanisms, and potential treatment enhancement is indis-
pensable. These issues are separately discussed in the following paragraphs.
Chronicity in anxiety disorders

A chronic course of anxiety disorders is estimated around 42%.3 More specifically, only
35% of patients with a social phobia recover during the 10 years after onset,4 40% of
patients with a panic disorder showed little improvement during a 8-12 year follow-up
period,5,6 30% of patients with GAD experienced chronic symptoms during a 2-year
follow-up,7 and 60% of the patients with OCD suffer from chronic symptoms.8 A clinical
problem with the interpretation of these chronicity rates is the lack of an unambiguous
definition of chronicity. The term chronicity has not only been used to describe the
persistence of a full anxiety disorder at follow-up, but it also serves to indicate the per-
sistence of residual symptoms (despite some remission). Moreover, in studies on chro-
nicity, follow-up periods differ and are often relatively short. Therefore, there is no con-
sensus between researchers and clinicians on when one should regard a disorder as
chronic. In addition, relapse or partial remission during the follow-up period are usually
ignored. Finally, it should be noted that chronicity and treatment resistance are over-
lapping but distinct phenomena. Chronic course refers to the persistence of symptoms
with or without adequate treatment, and treatment resistance refers to the persistence
of symptoms despite adequate treatment.
Whereas much research has been performed on the course and chronicity of
schizophrenia and depressive disorder, chronicity in anxiety disorders has been ne-
glected for a long time. The lack of attention to this topic is illustrated by the Dutch mul-
tidisciplinary guidelines for anxiety disorders in which the term ‘chronic’ (or ‘treatment
resistant’) is used only seven times (www.ggzrichtlijnen.nl). In contrast, in both the
Dutch multidisciplinary guidelines for depressive disorder and for schizophrenia, a full
chapter is dedicated to the subject of treatment resistance (www.ggzrichtlijnen.nl). The
deficiency of attention to chronicity in the anxiety disorders guidelines is a reflection of
the lack of longer-term follow-up studies until recently.
Fortunately, two recent large-scaled projects have been designed to investigate the
course of anxiety disorders and the risk factors and consequences of chronicity; these
are the Harvard/Brown Anxiety Research Program (HARP), and the Netherlands Study
of Depression and Anxiety (NESDA). The HARP study has followed over 700 patients
with an anxiety disorder during 10 years.9 The NESDA study encompasses a natural-
istic cohort study of 2,981 participants, which are recruited from general population,
general practices, and mental health organizations in the Netherlands.10
8
Introduction
OCD needs special attention in this introduction. Although categorised as an

anxiety disorder in the DSM-IV, both the HARP and the NESDA study did not include
patients with OCD. Therefore, research on the course of OCD is even more limited and
evidence to date is mainly derived from small, fragmented studies with methodological
shortcomings. The Netherlands Obsessive Compulsive Disorder Association (NOCDA)
study and the Brown Longitudinal Obsessive Compulsive Study were set up to address
this information gap. The NOCDA study entails an ongoing naturalistic longitudinal 1
cohort study conducted among 419 adult patients with a lifetime OCD,11 and the Brown
study included 293 patients with OCD in a prospective naturalistic study.12
Risk factor for complex anxiety disorders; comorbidity

Why do anxiety disorders run a relatively unfavourable course? The first issue con-
tributing to the unfavourable outcome of anxiety disorders is comorbidity. Comorbidity
in anxiety disorders is highly prevalent13 and an independent predictor of chronicity,3
whereas findings for other risk factors are less consistent. Comorbidity between anxi-
ety disorders and depressive disorders has been widely investigated and is associated
with childhood trauma, neuroticism,14 younger age of onset,15 illness severity, poorer
outcome,16-18 and impaired functioning.14,19
Although anxiety-anxiety comorbidity occurs as frequently as anxiety-depressive
comorbidity, it has received surprisingly little attention, as has comorbidity in OCD.
Moreover, literature on the relationship of comorbidity and complex anxiety disorders is
non-existent. Chronicity in anxiety disorders (including OCD) has been associated with
the presence of childhood trauma, high levels of neuroticism, severity and duration of
the primary disorder, and comorbidity.3,20-24
Since comorbidity studies are central in this thesis, several important aspects of
comorbidity are discussed in this paragraph. First, difficulties regarding the definition
of comorbidity are addressed. Second, epidemiologic data of comorbidity in anxiety
disorders are summarised. Next, issues regarding the clinical relevance of comorbidity
are discussed. Finally, comorbidity in the light of the DSM classification is discussed
combined with the current lack of biological substrate to define psychiatric entities.
Definition
In most literature, comorbidity is defined as the presence of two or more psychiatric
disorders according to DSM-IV criteria in one individual. This is, however, not in line
with the original definition of comorbidity, as introduced by Fernstein (1970).25 He de-
fined comorbidity in medicine as the co-existence of two distinct diseases. This small
difference in definitions has great impact, since distinguishing diseases in psychiatry
based on, for example neurobiological or genetic differences in etiology, has not been
very successful to date. More issues regarding this discrepancy are discussed later on
in this introduction.
Epidemiology
When the mostly used definition of comorbidity is used, the presence of two or more
DSM-IV psychiatric disorders, comorbidity is highly prevalent. Comorbidity within anx-
iety disorders and between anxiety and depressive disorders is rather common with
9
Chapter 1
current comorbidity rates of 40% for anxiety-anxiety comorbidity11,13,26 and 30% for
anxiety-depressive comorbidity.11,13,27 When comorbid disorders are present, anxiety
disorders become chronic in approximately 58% of the cases.7
Clinical relevance of comorbidity

Although these high comorbidity rates are recognised by clinicians, they are reluctant
to systematically address all comorbid DSM-IV diagnoses in their treatments.28,29 This
might be due to the idea that comorbidity is irrelevant in the treatment of the primary
disorder, which is supported by the lack of attention paid to comorbidity in the current
treatment guidelines of anxiety disorders. As mentioned before, the clinical implications
of comorbidity are not addressed in these guidelines, except for the co-occurrence of
major depressive disorder with an anxiety disorder. In case of this type of comorbidity,
the guidelines suggest pharmacological treatment in addition to or prior to psychologi-
cal treatment. For other types of comorbidity in anxiety disorders, the current opinion is
that comorbid disorders may disappear when the primary diagnosis is treated success-
fully30,31 and that caution is advised with non-specific treatment groups.32
Hence, it is unsurprising that clinicians are reluctant to address all DSM-IV diag-
nosis when they think it is irrelevant for the treatment. In this thesis, we investigate the
impact of comorbidity and hope to expand current knowledge about this topic. Further-
more, we aim to provide better clinical implications and guidelines for comorbidity. We
believe that when comorbidity is important for treatment, clinicians automatically will
improve the systematically diagnoses of all disorders present.
Comorbidity in the light of the DSM

Since comorbidity is common, yet not widely recognised as clinically relevant, it is
important to examine the nature and history of comorbidity in the light of the DSM. As
previously mentioned, the original definition of comorbidity is not useable in current
practice since the etiology of psychiatric disorders is largely unknown. The DSM is a
classification system based on categories, not on distinct diseases. Moreover, with ev-
ery new edition of the DSM, more categories are introduced, leading to higher comor-
bidity rates. For this reason, critics of the DSM classification system have addressed
psychiatric comorbidity as an artefact of this system and have argued that this type
of comorbidity is a correlate of severity of the primary disorder rather than the co-ex-
istence of two distinct diseases. Furthermore, anxiety and depressive disorder have
highly overlapping symptom criteria.
The authors of the DSM system have emphasised that DSM classifications must
not be seen as distinct entities and that comorbidity is indeed (partly) an artificial con-
sequence of the current classification.33 They also emphasise that this does not mean
that the DSM system fails. As long as research on genetic or neurobiological origins
fails to define distinct psychiatric diseases, a universal classification could be the best
alternative. In addition, the DSM system is ‘work in progress’ and thus it changes con-
stantly. For example, in the DSM-I and II, anxiety symptoms were described across the
full range of psychiatric disorders. In the DSM-III and IV, anxiety disorders were defined
and classified into several categories. With the introduction of the DSM-5, the reposi-
tioning of anxiety symptoms across the full range of DSM categories is reintroduced,
10
Introduction
next to the classification of anxiety disorders.2 For instance, anxiety distress has be-
come a specifier for depression in the DSM-5. Research has taken this repositioning a
step further with proposals to dimensionally assess panic attacks across the full range
of psychopathology34 and with the assumption that OC symptoms in psychosis might
be a valid specifier of psychosis since OC symptoms have been associated with worse
outcome for psychosis.35 Another direction in current research is the development
of models for profiling and staging of psychiatric disorders. These models cover the 1
range from presymptomatic stages up to severe, chronic stages,36-38 and comorbidity
assessment may substantially add to these models as staging specifiers of anxiety and
depressive disorders.
As mentioned above, one of the difficulties in defining distinct psychiatric diseases
is the lack of knowledge about the etiopathology of psychiatric disorders. Comorbid
disorders might reflect different manifestations of a single ‘core’ disease of which the
etiology is largely unknown.39 This view is supported by the fact that comorbidity rates
are high, treatment is similar,40 and comorbidity often diminishes when the primary
anxiety disorder is treated.41 However, research on etiopathology has not yet lead to
the description of distinct psychiatric diseases. Neuroimaging studies have shown
shared pathology in the anterior insula and dorsal anterior cingulate cortex across the
full range of psychopathology,42 but few disorder specific alterations. There is mounting
evidence that many DSM-IV defined distinct disorders (especially with respect to the
anxiety and depressive disorders) demonstrate substantial overlap of their underlying
neurobiological and genetic underpinnings. Therefore, many recent research initiatives
favour a transdiagnostic rather than a categorical disorder-based approach.43
This thesis
In sum, complex anxiety disorders tend to have a chronic course. An important risk
factor for chronicity, next to symptom severity, is the presence of a comorbid disorder.
Comorbidity between anxiety and depressive disorders is the rule rather than the ex-
ception. In addition, DSM classified anxiety and depressive disorders have highly over-
lapping symptom criteria and treatment options. These considerations suggest (par-
tially) shared underlying mechanisms. However, as long as psychiatric entities based
on underlying biological substrates are not yet known, comorbidity is defined as the
co-occurrence of two or more DSM-IV disorders. Despite the above-mentioned prob-
lems with this definition, this type of comorbidity might be relevant for clinical practice
as a specifier of disease severity and chronicity or to direct treatment.
As such, the relevance of comorbid disorders in anxiety disorders are investigated
in the first part of this thesis. In Chapter 2.1, the clinical relevance of type and number of
comorbid disorders in anxiety disorders is explored. A large sample (n=1004) of NESDA
participants with an anxiety disorder are divided into the following four groups: single
anxiety disorder, anxiety-anxiety comorbidity, anxiety-depressive comorbidity, and
“double” comorbidity (i.e. both anxiety and depressive comorbidity). These four groups
are compared according to sociodemographic-, vulnerability-, and clinical factors. In
Chapter 2.2, lifetime and current rates of comorbidity in a large sample of OCD patients
(n=382), drawn from the NOCDA sample, are examined in conjunction with the onset
and consequences of comorbidity in a large clinical sample of patients with OCD. A wide
11
Chapter 1
range of risk factors and clinical characteristics are also analysed to determine whether
‘pure’ OCD is different from OCD with current comorbidity. In Chapter 2.3, drawn from
the NESDA sample (n=2125), we investigate whether comorbid obsessive-compulsive
symptoms influence clinical outcomes in patients with anxiety and depressive disorders.
Underlying mechanism of complex anxiety disorders; expectancy bias

In this paragraph, we explain the relevance of expectancy bias regarding treatment out-
come in anxiety disorders. First, fear extinction mechanisms in animals and in humans
is briefly elucidated. Second, relevant literature identifying problems in fear acquisition
and extinction are described. Finally, the relevance of expectancy bias with regard to
the development and maintenance of anxiety disorders is addressed.
Investigating underlying mechanisms of anxiety disorders aids to understand why
anxiety disorders become chronic or complex and, ultimately, to improve treatment.
Fear acquisition and extinction

Theories and models for the development and treatment of anxiety disorders are
adapted from animal studies according to the concept of the classic ‘pavlovean’ fear
conditioning and extinction.44,45 In short, fear conditioning (for example, in rats) involves
an acquisition phase in which a neutral stimulus (for example, a light) is paired with an
aversive unconditioned stimulus (for example, a foot shock). The initially neutral stimu-
lus (the light) becomes a fear-provoking conditioned stimulus. In humans, a parallel, yet
not identical, process takes place. For instance, a shop in which someone once had
a panic attack and was frightened of having a heart attack may become a conditioned
stimulus for a panic disorder. After the first attack, the patient will associate the shop
with having a heart attack (even when such a panic attack does not occur), and, as a
potential consequence, the patient will avoid all shops.
In animal models of fear extinction, the fear-provoking conditioned stimulus (the
light) is repeatedly presented in absence of the unconditioned stimulus (the foot shock),
which usually results in extinction of the original fear response. This process is one
part of the beneficial effects of exposure therapy in anxiety disorders.46-48 Again, taking
panic disorder as a case in point, exposure to either panic symptoms (interoceptive
exposure) or to the shop (exposure in vivo) can lead to extinction of the automatic fear
response of having a heart attack. Nonetheless, it should be mentioned that fear ex-
tinction in humans is more complex and that this process of exposure is only one part
of fear extinction. Other factors (for instance, cognitions and expectations) also play a
key role in the development and treatment of anxiety disorders.
Problems in fear acquisition and extinction

Fear acquisition and extinction processes are physiological, adaptive phenomena.
Nevertheless, there are indications that anxiety patients display abnormalities in fear
acquisition and extinction that may be relevant for the onset of an anxiety disorder.
Patients with an anxiety disorder are found to have an enhanced fear acquisition and a
reduced fear extinction49 and an increased fear response to neutral stimuli during the
fear acquisition phase.50 These findings indicate that patients with an anxiety disorder
do not only have higher fear responses or reduced fear extinction, but might also have
12
Introduction
an impaired ability to inhibit fear (even in the presence of neutral stimuli, not condi-
tioned fear stimuli) and might have an increased tendency to generalise fear responses
to neutral stimuli resembling the conditioned fear stimuli.
Expectancy bias
Expectancy bias is the tendency to overestimate the association between fear-(ir)rel-
evant stimuli and the aversive consequence.51 Experimental research has established 1
that anxious individuals tend to overestimate this association. As an example, patients
with a panic disorder are exposed to fear relevant slides (a medical emergency room)
and fear irrelevant slides (a mushroom) that are possibly followed by a shock. Patients
tend to overestimate the contingency between the shock and both the fear-relevant
slides as well as the fear-irrelevant slides compared with healthy controls.52,53
Problems in expectancy bias may also play a key role in the maintenance of anxiety
disorders. For instance, one study has shown that the presence of expectancy bias in
spider fear predicted relapse two years later.54 This indicates that an expectancy bias
might be a predictor of chronicity in anxiety disorders.
This thesis
To summarise, patients with an anxiety disorder have stronger fear responses during
fear acquisition and extinction, and they tend to overestimate threat. This expectancy
bias may not only play a role in the development and maintenance of anxiety disorders,
but may also be relevant for treatment outcome. In Chapter 3, we assess whether in-
dividual differences in expectancy bias has direct predictive value regarding exposure
therapy outcome. A sample of 71 patients who have panic disorder with agoraphobia
are compared with 65 matched controls.
Treatment enhancement in anxiety disorders; d-cycloserine

In this paragraph, treatment issues in complex anxiety disorders are described. We first
shortly describe current treatment with its limitations and subsequently explicate possi-
ble treatment improvement using d-cycloserine as a treatment enhancer.
Current treatment
Current treatment for anxiety disorders involves several forms of psychotherapy (includ-
ing exposure therapy and cognitive behavioural therapy (CBT)), medication, or a com-
bination of both. CBT is effective, with 50-80% of the patients reporting clinical relevant
improvement55,56 and effect sizes around 0.7.57 However, there is still substantial room for
improvement since relapse and treatment refractory are estimated as high as 50%.58-60
Pharmacologic treatment includes mainly antidepressants such as the Selective
Serotonin Reuptake Inhibitors (SSRIs), Tricyclic Antidepressants (TCAs), and benzo-
diazepines. SSRIs are effective in the different anxiety disorders, with 60-80% of the
patients responding,61 effect sizes estimated between 0.4-0.7,62,63 and number needed
to treat of 5.64
The combination of psychological and pharmacological treatments has demon-
strated some advantages over monotherapy with an effect size of 0.43 and number
required to treat of 4.2.65
13
Chapter 1
Thus, current treatment options seem to be effective in at least part of the patients
with anxiety disorders, yet there is room for improvement since complex anxiety disor-
ders are associated with treatment resistance.
New developments
An interesting development in treatment improvement is based on so-called ‘transla-
tional psychiatry’. Instead of coincidently found medication (such as penicillin or benzo-
diazepines), translational psychiatry investigate pharmacological treatments based on
the knowledge of neurobiological underpinnings of anxiety disorders.
During the past decades, there is growing evidence for the role of glutamate in
stress response and anxiety. Chronic stress can lead to glutamate excitotoxicity, which
can lead to neurologic damage or cell death.66,67 Although it is estimated that around
80% of the synapses in the brain are glutamatergic, specific glutamatergic receptors
are involved in anxiety disorders. These N-methyl-D-aspartate (NMDA) receptors are
located in the amygdala, hippocampus, and prefrontal cortex (brain areas involved in
fear conditioning processes).68,69 Therefore, enhancing glutamatergic neurotransmis-
sion by influencing the NMDA receptor in prefrontal areas form the primary targets for
new drugs.70,71 This NMDA receptor is involved in the acquisition, consolidation, recon-
solidation, and extinction of fear memory72 and thus an interesting topic for translational
research. D-cycloserine (DCS) was known to be operant as a partial NMDA receptor
agonist and was hypothesised to be effective in enhancing extinction processes in the
treatment of anxiety disorders.
Furthermore, animal research has shown that DCS facilitates fear extinction learn-
ing73 and produces a generalised extinction effect74 when administrated before or after
extinction training in rodents.73,75 In addition, DCS might block post-extinction rein-
statement after a stressful event.76 Since extinction learning is considered as the core
mechanism of exposure therapy,68,77 these results suggest that DCS can provide the
following benefits for clinical treatment: enhancement of exposure therapy, a faster rate
of extinction of anxiety, and generalisation to cues that are not specifically addressed
in treatment. As a consequence, DCS augmentation might lead to a superior treatment
effect and to a lesser amount of exposure in vivo sessions.
Up until now, 13 randomised controlled trials have investigated the potential en-
hancing effect of DCS as an addition to several forms of psychotherapy in anxiety
disorders; these include 4 studies in OCD,78-81 2 in panic disorder with and without ago-
raphobia,82,83 2 in social phobia,84,85 2 in post-traumatic stress disorder (PTSD)86,87 and
3 in specific phobia.88-90 In these studies, the administered dosages of study medication
ranged between 50 mg and 250 mg, the number of sessions augmented with DCS
ranged between 1-12 sessions, and timing of administration varied (between 4 hours
before and directly after therapy).
Three meta-analyses were conducted on this topic.91-93 These meta-analyses found
effect sizes between 0.34-0.42. Moreover, DCS augmentation seemed to be most ef-
fective when administered for a limited number of times and at low dosages.
There is evidence from animal studies that DCS not only enhances fear extinction,
but that it enhances fear conditioning processes as well. This could indicate that the
use of DCS carries the risk of enhancing unsuccessful exposure sessions.94 Animal
14
Introduction
studies found DCS augmentation post extinction learning experiments to be equally

effective as directly preceding extinction learning,73,95 yet direct comparative studies in
humans are non-existing. Recently, host-hoc analyses of clinical data have revealed
that DCS enhancement of exposure therapy might only be effective in successful expo-
sure sessions.96,97
Even though DCS seems to improve extinction consolidation, the precise mecha-
nisms are still undetected. There are some unexplained differences in effect of DCS 1
between animal and human studies, as animal studies yield significantly larger effect
sizes (δ=1.19) versus studies performed in humans (δ=0.42).92 Grillon (2009)98 explains
these dissimilarities with the dual-model theory of fear conditioning. In this model, fear
extinction in animals is an automatic, low-level process, while extinction in humans
should also involve higher-order learning. Hence, it may be possible that DCS mainly
improves the lower level process and not the higher-level process. This automatically
addresses another difficulty in comparing human studies; every study uses a different
treatment protocol of CBT, exposure, or a combination. Another aspect of the unex-
plained differences between animal and human studies is the lack of knowledge on op-
timal timing, dosage, and frequency of administration of DCS. Due to the fact that DCS
was already proven safe, well-tolerated, and used in high dosages as anti-tuberculosis
medication, it was rapidly used in clinical studies; consequently, many pre-clinical tests
have not been conducted such as an adequate dose-effect study. As a result, optimal
timing (how long before or after the session), dosage (which dosage is the most effec-
tive), and frequency (how many sessions should optimally be augmented with DCS) is
not yet known.
This thesis
In Chapter 4, results are presented of our randomised, double-blinded, placebo-con-
trolled augmentation trial examining: i) the effectiveness of the addition of 125 mg DCS
to exposure therapy (before or directly after) in patients with panic disorder with agora-
phobia; and ii) the effectiveness of DCS augmentation prior to exposure compared with
DCS augmentation after exposure.
Aims of this thesis

This thesis aims to add to current knowledge about risk factors, underlying mecha-
nisms, and treatment of (complex) anxiety disorders. The objective of the first part of
the thesis is to investigate the clinical relevance of comorbidity in anxiety disorders. In
Chapter 2.1, the relevance of type and number of comorbid disorders in anxiety disor-
ders is investigated. Chapter 2.2 explores the relevance of comorbidity OCD patients,
and in chapter 2.3, the relevance of comorbid obsessive-compulsive symptoms in anxi-
ety and depressive disorders are examined.
The second part of this thesis focuses on a possible underlying mechanism of
anxiety disorders. In Chapter 3, we investigate the presence of expectancy bias as a
predictor of exposure therapy outcome.
The third part of this thesis concentrates on a new treatment enhancer called d-cy-
closerine (DCS). The RCT investigating the effectiveness of adding DCS to exposure
therapy in patients with panic disorder with agoraphobia are discussed here.
15
Chapter 1
Finally, in Chapter 5, a summary of the main results is provided and discussed in a

broader perspective.
16
Introduction
Reference list
1. Bijl RV, Ravelli A, van Zessen G. Prevalence of psychiatric disorder in the general
population: results of The Netherlands Mental Health Survey and Incidence Study
(NEMESIS). Soc Psychiatry Psychiatr Epidemiol 1998;33:587-595.
2. American Psychiatric Association. Diagnostic and Statistical Manual of Mental Dis-
orders, 4th edn. Washington DC: American Psychiatric Association;1994.
3. Penninx BW, Nolen WA, Lamers F et al. Two-year course of depressive and anx- 1
iety disorders: Results from the Netherlands Study of Depression and Anxiety
(NESDA). Journal of Affective Disorders 2011;33:76–85.
4. Keller MB. Social anxiety disorder clinical course and outcome: review of Harvard/
Brown Anxiety Research Project (HARP) findings. J Clin Psychiatry 2006;67 Suppl
12:14-19.
5. Keller MB, Hanks DL. Course and outcome in panic disorder. Prog Neuropsycho-
pharmacol Biol Psychiatry 1993;17:551-570.
6. Yonkers KA, Bruce SE, Dyck, IR et al. Chronicity, relapse, and illness course of
panic disorder, social phobia, and generalized anxiety disorder: findings in men
and women from 8 years of follow-up. Depressi Anxiety 2003;17:173–179.
7. Hendriks SM, Spijker J, Licht CM et al. Two-year course of anxiety disorders: dif-
ferent across disorders or dimensions? Acta Psychiatr Scand 2013;128:212-221.
8. Visser HA, van Oppen P, van Megen HJ et al. Obsessive-compulsive disorder;
chronic versus non-chronic symptoms. J Affect Disord 2014;152-154:169-174.
9. Keller MB, Yonkers KA, Warshaw MG et al. Remission and relapse in subjects with
panic disorder and panic with agoraphobia: a prospective short interval naturalistic
follow-up. J Nerv Ment Dis 1994;182:290–296.
10. Penninx BW, Beekman AT, Smit JH et al. The Netherlands Study of Depression
and Anxiety (NESDA): Rationale, Objectives and Methods. Int J Methods Psychiatr
Res 2008;17:121-140.
11. Schuurmans J, van Balkom AJ, van Megen HJ et al. The Netherlands OCD Asso-
ciation study: Design and rationale of a longitudinal naturalistic study of the course
of OCD and clinical characteristics of the sample at baseline. Int J Methods Psy-
chiatr Res 2012;21:273-85.
12. Pinto A, Mancebo MC, Eisen JL et al. The Brown longitudinal obsessive compul-
sive study: Clinical features and symptoms of the sample at intake. J Clin Psychia-
try 2006:67;703-711.
13. Brown TA, Campbell LA, Lehman CL et al. Current and lifetime comorbidity of the
DSM-IV anxiety and mood disorders in a large clinical sample. J Abnorm Psychol
2001;110:585-599.
14. de Graaf R, Bijl RV, Smit F et al. Risk factors for 12-month comorbidity of mood,
anxiety, and substance use disorders: findings from the Netherlands Mental Health
Survey and Incidence Study. Am J Psychiatry 2002;159:620-629.
15. Fichter MM, Quadflieg N, Fischer UC et al. Twenty-five-year course and outcome
in anxiety and depression in the Upper Bavarian Longitudinal Community Study.
Acta Psychiatr Scand 2010;122:75-85.
16. Angst J, Vollrath M. The natural history of anxiety disorders. Acta Psychiatr Scand
1991;84:446-452.
17
Chapter 1
17. Merikangas KE, Zhang H, Avenevoli S et al. Longitudinal Trajectories of Depres-

sion and Anxiety in a Prospective Community Study. The Zurich Cohort Study.
Arch Gen Psych 2003;60:993-1000.
18. Rush AJ, Zimmerman M, Wisniewski SR et al. Comorbid psychiatric disorders
in depressed outpatients: Demographic and clinical features. J Affect Disord
2005;87:43-55.
19. Kessler RC, Frank RG. The impact of psychiatric disorders on work loss days. Psy-
chol Med 1997;27:861-873.
20. Bruce SE, Yonkers KA, Otto MW et al. Influence of psychiatric comorbidity on
recovery and recurrence in generalized anxiety disorder, social phobia, and panic
disorder: a 12-year prospective study. Am J Psychiatry 2005;162:1179-87.
21. Gibb BE, Chelminski I, Zimmerman M. Childhood emotional, physical, and sexual
abuse, and diagnoses of depressive and anxiety disorders in adult psychiatric out-
patients. Depress Anxiety 2007;24:256-263.
22. Scholten WD, Batelaan NM, van Balkom AJ et al. Recurrence of anxiety disorders
and its predictors. Journal of affective disorders 2013;147:180-185.
23. Selvi Y, Besiroglu L, Aydin A et al. Relations between childhood traumatic experi-
ences, dissociation, and cognitive models in obsessive compulsive disorder. Int J
Psychiatry Clin Pract 2012;16:53–59.
24. Visser HA, van Minnen A, van Megen HJ et al. The relationship between adverse
childhood experiences and symptom severity, chronicity, and comorbidity in pa-
tients with obsessive-compulsive disorder. J Clin Psychiatry 2014;75:1034-1039.
25. Fernstein AR. The pre-therapeutic classification of co-morbidity in chronic diseas-
es. J Chronic Dis 1970;23:455–468.
26. Lamers F, Van Oppen P, Comijs HC et al. Comorbidity patterns of anxiety and de-
pressive disorders in a large cohort study. J Clin Psychiatry 2011;72:341-348.
27. Denys D, Tenney N, van Megen HJ et al. Axis I and II comorbidity in a large sam-
ple of patients with obsessive-compulsive disorder. J Affect Disord 2004;80:155-
162.
28. Ramirez Basco M, Bostic JQ, Davies D et al. Methods to improve diagnostic accu-
racy in a community mental health setting. Am J Psychiatry 2000;157:1599-1605.
29. Zimmerman M, Mattia JI. Psychiatric diagnosis in clinical practice: is comorbidity
being missed? Compr Psychiatry 1999;40:182-191.
30. Borkovec TD, Abel J L, Newman H. Effects of psychotherapy on comorbid condi-
tions in generalized anxiety disorder. J Consult Clin Psychol 1995;63:479–483.
31. Tsao JC, Mystkowski JL, Zucker BG, Craske MG. Effects of cognitive-behavioral
therapy for panic disorder on comorbid conditions: Replication and extension. Be-
hav Res Ther 2002;33:493–509.
32. Craske MG. The cutting edge: Transdiagnostic treatment for anxiety and depres-
sion. Depress Anxiety 2012;29:749-753.
33. First MB. Mutually exclusive versus co-occurring diagnostic categories: the chal-
lenge of diagnostic comorbidity. Psychopathology 2005;38:206-210.
34. Batelaan NM, Rhebergen D, de Graaf R et al. Panic Attacks as s Dimension of
Psychopathology: Evidence for Associations With Onset and Course of Mental
Disorders and Level of Functioning. J Clin Psychiatry 2012;73:1195-1202.
18
Introduction
35. Van Dael F, van Os J, de Graaf R et al. Can Obsessions Drive You Mad? Longi-
tudinal Evidence that Obsessive-Compulsive Symptoms Worsen the Outcome of
Early Psychotic Experiences. Acta Psychiatr Scand 2011;123:136-146.
36. Balkom van AJ, Oosterbaan DB, Batelaan N et al. The characterization of anxiety
disorders: staging and profiling based on common sense. Tijdschrift voor psychi-
atrie 2012;11:935-940.
37. Fava GA, Tossani E. Prodromal stage of major depression. Early Intervention in 1
Psychiatry 2007;1: 9-18.
38. Fava GA, Rafanelli C, Tossani E, Grandi S. Agoraphobia is a disease: a tribute to
Sir Martin Roth. Psychother Psychosom 2008;77:133-138.
39. Maj M. The aftermath of the concept of ‘psychiatric comorbidity’. Psychother Psy-
chosom 2005;74:67-68.
40. Ressler KJ, Mayberg HS. Targeting abnormal neural circuits in mood and anxiety
disorders: from the laboratory to the clinic. Nat Neurosci 2007;10:1116-1124.
41. Brown TA, Leyfer O. Classification of Anxiety Disorders. In D.J.Stein, E.Hollander,
& B.O. Rothbaum (Eds.), Textbook of Anxiety Disorders (2nd ed., pp. 17-35). Amer-
ican Psychiatric Publishing, Inc. 2009.
42. Goodkind M, Eickhoff SB, Oathes DJ et al. Identification of a Common Neurobio-
logical Substrate for Mental Illness. JAMA Psychiatry 2015;72:305-315.
43. Cuthbert BN. The RDoC framework: facilitating transition from ICD/DSM to dimen-
sional approaches that integrate neuroscience and psychopathology. World Psy-
chiatry 2014;13:28-35.
44. Mineka S, Oehlberg K. The relevance of recent developments in classical condi-
tioning to understanding the etiology and maintenance of anxiety disorders. Acta
Psychol 2008;127:567-580.
45. Pavlov IP. Conditioned reflexes London: Oxford University Press; 1927.
46. Hofmann SG. Cognitive processes during fear acquisition and extinction in ani-
mals and humans: Implications for exposure therapy of anxiety disorders. Clinical
Psychology Review 2008;28:199-210.
47. Massad PM, Hulsey TL. Exposure therapy renewed. Journal of Psychotherapy In-
tegration 2006;16:417-428.
48. Myers KM, Davis M. Chapter 2.3 extinction of fear: From animal studies to clinical
interventions. Handbook of behavioral neuroscience (pp. 49-62) Elsevier; 2008.
49. Lissek S, Powers AS, McClure EB et al. Classical fear conditioningin the anxiety
disorders:a meta-analysis. Behaviour Research and Therapy 2005;43:1391–1424.
50. Duits P, Cath DC, Lissek, S et al. Updated meta-analysis of classical fear condi-
tioning in the anxiety disorders. Depress Anxiety 2015;32:239-253.
51. Davey GC. An expectancy model of laboratory preparedness effects. Journal of
Experimental Psychology 1992;121:24-40.
52. de Jong PJ, Merckelbach H, Arntz A, Nijman H. Covariation detection in treated
and untreated spider phobics. Journal of Abnormal Psychology 1992;101:724-727.
53. Wiedemann G, Pauli P, Dengler W. A priori expectancy bias in patients with panic
disorder. Journal of Anxiety Disorders 2001;15:401-412.
54. de Jong PJ, van den Hout MA, Merckelbach H. Covariation bias and the return of
fear. Behaviour Research and Therapy 1995;33:211-213.
19
Chapter 1
55. Barlow DH, Gorman JM, Shear MK et al. Cognitive behavioral therapy, imipramine
or their combination for panic disorder. JAMA 2000;283;2529-2536.
56. Allen LB, White KS, David H. Barlow DH et al. Cognitive-Behavior Therapy (CBT)
for Panic Disorder: Relationship of Anxiety and Depression Comorbidity with Treat-
ment Outcome. J Psychopathol Behav Assess 2010;32:185–192.
57. Hofmann SG, Smits JA. Cognitive-behavioral therapy for adult anxiety disor-
ders: a meta-analysis of randomized placebo-controlled trials. J Clin Psychiatry
2008;69:621–632.
58. Durham RC, Chambers JA, Power KG et al. Long-term outcome of cognitive be-
haviour therapy clinical trials in central Scotland. Health Technol Assess 2005;9:1-
174.
59. Rachman S. The return of fear: Review and prospect. Clinical Psychology Review
1989;9:147-168.
60. Ravindran LN, Stein MB. The pharmacologic treatment of anxiety disorders: a re-
view of progress. J Clin Psychiatry 2010;71:839-854.
61. Hoffman EJ, Mathew SJ. Anxiety Disorders: A Comprehensive Review of Pharma-
cotherapies. Mount Sanai Journal of Medicine 2008;75:248–262.
62. Blanco C, Schneier FR, Schmidt A et al. Pharmacological treatment of social anxi-
ety disorder: a meta-analysis. Depress Anxiety 2003;18:29–40.
63. Otto MW, Tuby KS, Gould RA et al. An effect-size analysis of the relative efficacy
and tolerability of serotonin selective reuptake inhibitors for panic disorder. Am J
Psychiatry 2001;158:1989–1992.
64. Kapczinski F, Lima MS, Souza JS, Schmitt R. Antidepressants for generalized
anxiety disorder. Cochrane Database Syst Rev 2003;2:CD003592.
65. Cuijpers P, Sijbrandij M, Koole SL et al. Adding psychotherapy to antidepressant
medication in depression and anxiety disorders: a meta-analysis. World Psychiatry
2014;13:56-67.
66. Kendell SF, Krystal JH, Sanacora G. GABA and glutamate systems as therapeutic
targets in depression and mood disorders. Expert opinion on therapeutic targets
2005;9:153-168.
67. Cortese BM, Phan KL. The role of glutamate in anxiety and related disorders. CNS
spectrums 2005;10:820-830.
68. Quirk GJ, Garcia R, Gonzalez-Lima F. Prefrontal mechanisms in extinction of con-
ditioned fear. Biology Psychiatry 2006;60:337-343.
69. Gillepsie CF, Ressler KJ. Emotional learning and glutamate: translational perspec-
tive. CNS spectrums 2007;10:831-839.
70. Miserendino MJ, Sananes CB, Melia KR, Davis M. Blocking of acquisition but not
expression of conditioned fear-potentiated startle by NMDA antagonists in the
amygdala. Nature 1990;345:716–718.
71. Walker DL, Davis M. The role of amygdala glutamate receptors in fear learning,
fear-potentiated startle, and extinction. Pharmacol Biochem Behav 2002;71:379–
392.
72. Amaral OB, Roesler R. Targeting the NMDA receptor for fear-related disorders.
Recent Pat CNS Drug Discov 2008;3:166-178.
73. Ledgerwood L, Richardson R, Cranney J. Effects of D-cycloserine on extinction of
20
Introduction
conditioned freezing. Behav. Neurosci 2003;117:341-349.

74. Ledgerwood L, Richardson R, Cranney J. D-cycloserine facilitates extinction of
learned fear: effects on reacquisition and generalized extinction. Biol. Psychiatry
2005;57:841-847.
75. Walker DL, Ressler KJ, Lu KT, Davis M. Facilitation of conditioned fear extinction
by systemic administration or intra-amygdala infusions of D-cycloserine as as-
sessed with fear-potentiated startle in rats. J Neurosci 2002;22:2343-2351. 1
76. Ledgerwood L, Richardson R, Cranney J. D-cycloserine and the facilitation of
extinction of conditioned fear: consequences for reinstatement. Behav Neurosci
2004;118:505-513.
77. Bouton ME. A learning theory perspective on lapse, relapse, and the maintenance
of behavior change.Health Psychol 2000;19:57-63.
78. Kushner MG, Kim SW, Donahue C et al. D-cycloserine augmented exposure thera-
py for obsessive-compulsive disorder. Biol Psychiatry 2007;62:835-838.
79. Storch EA, Merlo LJ, Bengtson M et al. D-cycloserine does not enhance expo-
sure-response prevention therapy in obsessive-compulsive disorder. Int Clin Psy-
chopharmacol 2007;22:230-237.
80. Storch EA, Murphy TK, Goodman WK et al. A preliminary study of D-cycloserine
augmentation of cognitive-behavioral therapy in pediatric obsessive-compulsive
disorder. Biol Psychiatry 2010;68:1073-1076.
81. Wilhelm S, Buhlmann U, Tolin DF et al. Augmentation of behavior therapy with
D-cycloserine for obsessive-compulsive disorder. Am J Psychiatry 2008;165:335-
341.
82. Otto MW, Tolin DF, Simon NM et al. Efficacy of D-cycloserine for enhanc-
ing response to cognitive-behavior therapy for panic disorder. Biol Psychiatry
2010;67:365-370.
83. Siegmund A, Golfels F, Finck C et al. D-Cycloserine does not improve but might
slightly speed up the outcome of in-vivo exposure therapy I patients with severe
agoraphobia and panic disorder in a randomized double blind clinical trial. J Psy-
chiatr Res 2011;45:1042-1047.
84. Guastella AJ, Richardson R, Lovibond PF et al. A randomized controlled trial of
D-Cycloserine Enhancement of Exposure therapy of social anxiety disorder. Biol
Psychiatry 2008;63:544-549.
85. Hofmann SG, Meuret AE, Smits JA et al. Augmentation of exposure therapy with
D-cycloserine for social anxiety disorder. Arch Gen Psychiatry 2006;63:298-304.
86. de Kleine RA, Hendriks G, Kusters WJ et al. A randomized placebo-controlled trial
of d-cycloserine to enhance exposure therapy for posttraumatic stress disorder.
Biol Psychiatry 2012;11:962-968.
87. Litz BT, Salters-Pedneault K, Steenkamp MM et al. A randomized placebo con-
trolled trial of D-cycloserine and exposure therapy for posttraumatic stress disor-
der. J Psychiatr Res 2012;46:1184-1190.
88. Nave AM, Tolin DF, Stevens MC et al. Exposure Therapy, D-Cycloserine, and
Functional Magnetic Resonance Imaging in Patients With Snake Phobia: A Ran-
domized Pilot Study. J Clin Psychiatry 2012;73:1179-1186.
89. Ressler KJ, Rothbaum BO, Tannenbaum L et al. Cognitive enhancers as adjuncts
21
Chapter 1
to psychotherapy: use of D-cycloserine in phobic individuals to facilitate extinction

of fear. Arch Gen Psychiatry 2004;61:1136-1144.
90. Tart CD, Handelsman PR, De Boer LB et al. Augmentation of exposure therapy
with post-session administration of D-cycloserine. J Psychiatr Res 2013;47:168-
174.
91. Bontempo A, Panza KE, Bloch MH et al. D-cycloserine augmentation of behavioral
therapy for the treatment of anxiety disorders: a meta-analysis. J Clin Psychiatry
2012;73:533-537.
92. Norberg, MM, Krystal JH, Tolin DF et al. A Meta-Analysis of D-cycloserine and the
facilitation of fear extinction and exposure therapy. Biol Psychiatry. 2008;63:1118-
1126.
93. Rodrigues H, Figueira I, Lopes A et al. Does D-Cycloserine Enhance expo-
sure therapy for anxiety disorders in humans? A meta-analysis. PLoS ONE.
2014;9:e93519.
94. Lee JL, Milton AL, Everitt BJ. Reconsolidation and extinction of conditioned fear:
inhibition and potentiation. J Neurosci 2006;26:10051-10056.
95. Parnas AS, Weber M, Richardson R. Effects of multiple exposures to D-cycloser-
ine on extinction of conditioned fear in rats. Neurobiol Learn Mem 2005;83:224-
231.
96. Smits JA, Rosenfield D, Otto MW, et al. D-cycloserine enhancement of exposure
therapy for social anxiety disorder depends on the success of exposure sessions.
J Psychiatr Res. 2013;47:1455-1461.
97. Smits Ja, Rosenfield D., Otto MW, et al. D-cycloserine enhancement of fear Ex-
tinction is specific to succesfull exposure sessions: Evidence from the treatment of
height phobia. Biol psychiatry. 2013;73:1054-1058.
98. Grillon. D-Cycloserine Facilitation of Fear Extinction and Exposure-Based Therapy
Might Rely on Lower-Level, Automatic Mechanisms. Biol psychiatry 2009;66:636–
641.
22
Introduction
23
24
Chapter 2.1
Clinical relevance of comorbidity in anxiety disorders:

A report from the Netherlands Study of Depression and
Anxiety (NESDA)
Mieke Klein Hofmeijer-Sevink, Neeltje M. Batelaan, Harold J.G.M. van Megen,

Brenda W. Penninx, Danielle C. Cath, Marcel A. van den Hout, Anton J.L.M. van
Balkom.
Journal of Affective Disorders 2012;137:106-112.
Author contributions: M. Klein Hofmeijer-Sevink and A. van Balkom designed the study and wrote the proto-
col. M. Klein Hofmeijer-Sevink and N. Batelaan undertook the statistical analysis and wrote the first draft of the
manuscript. B. Penninx supervised the design of the study and the statistical analysis. H. van Megen, D. Cath
and M. van der Hout corrected the first draft of the manuscript. All authors contributed to and have approved
the final manuscript.
25
Chapter 2.1
Abstract
Background: To study the clinical relevance of type of comorbidity and number of co-
morbid disorders in anxiety disorders. Four groups were compared according to socio-
demographic-, vulnerability- and clinical factors: single anxiety disorder, anxiety-anxi-
ety comorbidity, anxiety-depressive comorbidity and “double” comorbidity (i.e. anxiety
and depressive comorbidity).
Methods: Data were obtained from the Netherlands Study of Anxiety and Depression
(NESDA). A sample of 1004 participants with a current anxiety disorder was evaluated.
Results: As compared with single anxiety, anxiety-anxiety comorbidity was associated
with higher severity, greater chronicity and more treatment. Anxiety-anxiety comorbidity
was associated with an earlier age of onset and a more chronic course compared with
anxiety-depressive comorbidity, while anxiety-depressive comorbidity was associated
with more severe symptoms and more impaired functioning than anxiety-anxiety co-
morbidity. “Double” comorbidity was associated with higher severity, greater chronicity,
more treatment and increased disability. Sociodemographic and vulnerability factors
were comparable among the four groups.
Limitations: A prospective design would be more appropriate to study the outcome.
In this study no distinction was made between whether depression or anxiety disorder
preceded the current anxiety disorder.
Conclusions: It is clinical relevant to diagnose and treat comorbidity among anxiety
disorders as it is associated with higher severity and more chronicity. Whereas anxi-
ety-anxiety comorbidity has an earlier age of onset and a more chronic course, anxi-
ety-depressive comorbidity leads to more treatment and impaired functioning. “Double”
comorbidity leads to even more severity, chronicity and impairment functioning com-
pared with both anxiety-anxiety and anxiety-depressive comorbidity.
26
Comorbidity in anxiety disorders
INTRODUCTION
Comorbidity in anxiety disorders represents the rule rather than the exception. Lifetime
comorbidity in patients with anxiety disorders occurs in more than 80%.1 The preva-
lence of having a current comorbid axis I disorder is estimated at 50%, with anxiety
disorders and depressive disorders being the most prevalent comorbid disorders.1,2
Comorbidity between anxiety disorders and depressive disorders has been widely
investigated. Anxiety-depressive comorbidity has been associated with sociodemo-
graphic factors such as female gender, not having a partner, lower socioeconomic sta-
tus and lower educational level;3-5 with vulnerability factors such as parental psychiatric
history, childhood trauma, negative life events and neuroticism;4 and with clinical fac- 2.1
tors including younger age of onset,3 illness severity, poorer outcome,6-8 higher health-
care utilization9 and impaired functioning.3,10
Although anxiety-anxiety comorbidity occurs as frequently as anxiety-depressive
comorbidity, it has received surprisingly little attention to-date. Fragmentized research
suggests that, compared with single anxiety disorders, patients with anxiety-anxiety
comorbidity suffer from more severe anxiety symptoms.11 In addition, anxiety-anxiety
comorbidity may be associated with lower levels of extraversion and higher levels of
neuroticism.12 Finally, individuals with anxiety-anxiety comorbidity more frequently re-
port a history of childhood abuse.13 However, anxiety-anxiety comorbidity may not be
associated with lower levels of functioning compared with single anxiety disorders.14-16
Given the lack of data regarding anxiety-anxiety comorbidity, further research is
needed. It is important to determine whether it is clinically relevant to diagnose type of
comorbidity and number of comorbid disorders in anxiety disorders so as to be able to
identify patients at a high risk for developing a comorbid disorder and to direct treat-
ment at the disorders present.
Aims of the study

Using data from the Netherlands Study of Anxiety and Depression (NESDA), the clinical
relevance of (i) type of comorbidity and (ii) number of comorbid disorders in a clinical
sample suffering from anxiety disorders was studied. Four groups were compared: sin-
gle anxiety disorder, anxiety-anxiety comorbidity, anxiety-depressive comorbidity and
“double” comorbidity (i.e. anxiety and depressive comorbidity). We investigated socio-
demographics, vulnerability factors and clinical characteristics to determine whether:
1) anxiety-anxiety comorbidity differs from a single anxiety disorder

2) anxiety-anxiety comorbidity differs from anxiety-depressive comorbidity
(type of comorbidity)
3) “double” comorbidity differs from anxiety-anxiety comorbidity and from
anxiety-depressive comorbidity (number of comorbid disorders).
METHODS
Procedure
Data were obtained from the Netherlands Study of Depression and Anxiety (NESDA).
NESDA is a longitudinal cohort study including 2981 adults (18-65 years) with anxiety
disorders (panic disorder with or without agoraphobia, social phobia, generalized anxi-
27
Chapter 2.1
ety disorder (GAD) and agoraphobia without a history of panic disorder) and/or depres-
sive disorders (major depressive disorder, dysthymic disorder) and healthy controls. A
detailed description of the NESDA study design can be found elsewhere.17 In short: the
NESDA study aims to describe the long-term course and consequences of depressive
and anxiety disorders in a sample recruited from the community, primary care settings
and specialized mental health care facilities. Specific phobias, post-traumatic stress
disorders and obsessive-compulsive disorders were not assessed and were therefore
not included in this study. Approval of the study was granted by the Ethical Review
Boards of all the participating centers and written informed consent was obtained from
all participants. For the present study we used data from the baseline interview, which
included a 4-hour interview, a medical assessment, computer tasks and two self-ad-
ministered questionnaires, all instruments were in Dutch.
Sample
Of the 2981 participants in the baseline interview, 1363 had had at least one 12-month
anxiety disorder. To avoid the confounding of comorbid alcohol dependence, 279 par-
ticipants with a 12-month diagnosis of alcohol dependence were excluded. Since our
research questions regard both type and number of comorbid disorders, we formed
four mutually exclusive groups based on those two aspects: participants with a single
anxiety disorder, without a depressive disorder (29.4%, n=295), with two anxiety dis-
orders, without a depressive disorder (10.5%, n=105), with one anxiety disorder and
one depressive disorder (29.1%, n=292) and with “double” comorbidity (i.e. two anxiety
disorders and one or two depressive disorders, 31.1%, n=312). In addition, 18 patients
with three anxiety disorders were excluded, as well as 62 with one anxiety disorder and
two depressive disorders. This resulted in a total sample size of 1004.
Assessment of psychopathology
To diagnose DSM-IV anxiety disorders and depressive disorders in the past 12 months,
the Dutch version of the Composite International Diagnostic Interview (CIDI, version
2.1) was used as a baseline assessment. The CIDI is a valid and reliable instrument for
axis I diagnoses.18,19 The CIDI was conducted by trained research staff.
Correlates of comorbidity
Sociodemographic factors included gender (male/female), age (in years), education (in
years) and partner status (present/absent).
Vulnerability factors included a family history of anxiety and depressive disorders,

childhood trauma, negative life events and personality traits. A first-degree family his-
tory of anxiety and depressive disorders was assessed using the family tree method.20
Childhood trauma was assessed using the second section of the Childhood Trauma
Interview previously used in the Netherlands Mental Health Survey and Incidence
Study (NEMESIS).4 Childhood trauma was considered present in cases when partici-
pants experienced emotional neglect, psychological abuse or physical abuse on two
or more occasions prior to age 16, or sexual abuse on one or more occasions prior to
age 16. Negative life events in the past year were measured with the List of Threaten-
28
ing Experiences.21 This list has good reliability and good validity.22 Finally, personality
traits were measured with the 60-item NEO-FFI personality questionnaire, which uses
5 dimensions for personality: neuroticism, extraversion, openness, agreeableness and
conscientiousness.23 This scale has proven to be reliable and valid.24
Clinical characteristics included age of onset (in years), chronicity of symptoms (yes/
no), severity of depressive, anxiety and avoidance symptoms (assessed as a contin-
uous variable), treatment (yes/no) and disability (assessed as a continuous variable).
Age of onset was determined with the CIDI, and the youngest age of onset was used
in the case of comorbidity. The chronicity of symptoms was based on the Life Chart 2.1
Interview25 in which the presence of anxiety and/or depressive symptoms was retro-
spectively assessed for the preceding 4 years. Symptoms were defined as chronic
when they occurred during more than 50% of the months during the preceding 4 years.
The Life Chart Interview has shown high validity and reliability.26 Receiving treatment
was defined as having received antidepressant medication and/or psychotherapy.
The use of antidepressant medication was determined on the basis of drug container
inspection for all drugs used in the month before assessment. Use of antidepressants
was considered positive when taken regularly (defined as more than 50% of the time/
daily for at least 1 month); these included TCAs, SSRIs and SNRIs. Psychotherapy was
defined as more than 5 contacts with a therapist during the past 6 months. A therapist
could be a psychologist, psychiatrist, psychotherapist, social worker or social psychiat-
ric nurse working in general health care or in specialized mental health care. The fre-
quency of contacts with a therapist was assessed using the Tic-P.27 This is a validated
instrument assessing loss of productivity at work and health care utilization. The se-
verity of anxiety symptoms was assessed with the Beck Anxiety Inventory (BAI).28 The
BAI is a 21-item self-report list. It is widely used and has proven to be highly valid and
reliable.29 The level of avoidance was measured with the Fear Questionnaire (FQ),30
a 15-item self-report list measuring phobic symptoms that has been proven valid in a
Dutch population.31 To determine the severity of depressive symptoms the Inventory of
Depressive Symptomatology (IDS)32 was used. This scale is a 30-item self-report list
measuring depressive symptoms that has proven to be valid and reliable.33 Negative
consequences in three domains (cognitive, physical and social disability) were mea-
sured with The World Health Organization Disability Assessment Schedule (WHODAS
II). This is a 36-item instrument that examines disability in six domains of life during the
past 30 days and has proven to be valid and reliable.34
Statistical analyses
All statistical analyses were conducted with the SPSS statistical package version 15.0.
As described above, the sample was divided into four mutually exclusive categories:
participants with a single anxiety disorder, with anxiety-anxiety comorbidity, with anxi-
ety-depressive comorbidity and with “double” comorbidity. Bivariate multinomial logistic
regression analyses were conducted to examine the clinical factors of comorbidity. De-
pendent on the research question, different groups were taken as the reference group.
Subsequently, risk-indicators (sociodemographics and vulnerability factors) were ex-
amined. Both bivariate and multivariate multinomial logistic regression analyses were
29
Chapter 2.1
conducted. In the multivariate model, we adjusted for all covariates that were significant
in the bivariate analyses.
RESULTS
Characteristics of the sample
The sociodemographics and clinical characteristics of the sample are presented in Ta-
ble 1. The sample was predominantly female (72.1%). The mean age of onset was 20.9
years (SD: 12.3). Forty-five percent of the sample suffered from chronic symptoms and
about half of the sample had received treatment (51.3%).
Table 1
Socio-demographic- and clinical characteristics of the sample (n=1004).
% (n)
Female gender 72.1 (724)
Partner 66.7 (670)
Chronicity 45.3 (455)
Treatment 51.3 (515)
Mean (SD)
Age (years) 41.0 (12.6)
Education (years) 11.7 (3.2)
Age of onset (years) 20.9 (12.3)
Total IDS score 28.3 (12.6)
Total BAI score 17.7 (10.7)
Total FQ score 34.7 (20.1)
Cognitive disability 31.5 (20.9)
Physical disability 18.7 (21.7)
Social disability 32.8 (20.6)
IDS= Inventory of Depressive Symptomatology, BAI= Beck Anxiety Inventory, FQ= Fear Questionnaire
Anxiety-anxiety comorbidity
To investigate whether anxiety-anxiety comorbidity is associated with various clinical
factors and various risk indicators compared with single anxiety disorders, participants
with a single anxiety disorder (n=295) were compared with those with anxiety-anxiety
30
comorbidity (n=105). The results are presented in Tables 2 and 3.

Participants with anxiety-anxiety comorbidity had an earlier age of onset than those
with a single anxiety disorder (19.3 versus 23.0 years; OR: 0.73; 95%CI: 0.58-0.93) and
a higher rate of chronicity (53.3% versus 32.2%; OR: 2.41; 95%CI: 1.53-3.79). Partici-
pants with anxiety-anxiety comorbidity experienced more severe depressive (OR: 1.52;
95%CI: 1.14-2.03), anxiety (OR: 1.58; 95%CI: 1.20-2.07) and avoidance symptoms (OR:
1.42; 95%CI: 1.10-1.82) than those with a single anxiety disorder. They experienced
more social disability compared with participants with a single anxiety disorder (OR:
1.45; 95%CI: 1.16-1.98), but similar rates of cognitive and physical disability. Partic-
ipants with anxiety-anxiety comorbidity did not receive significantly more treatment 2.1
(40.0% versus 30.8%; OR: 1.50; 95%CI: 0.94-2.37).
Unadjusted bivariate analyses show that participants with anxiety-anxiety co-
morbidity more often had a positive family history of anxiety or depressive disorders
(63.8% versus 46.4%) compared with participants with a single anxiety disorder. Those
with anxiety-anxiety comorbidity had higher neuroticism levels and lower extraversion
levels. The groups did not differ critically with respect to age at the baseline interview,
childhood trauma, negative life events in the past year and personality traits other than
neuroticism and extraversion. (Unadjusted statistics are not shown in the Table).
After adjusting for all covariates that were significant in the bivariate analyses, the
higher rate of positive family history (OR: 2.07; 95%CI: 1.30-3.30) and the elevated
neuroticism levels (OR: 1.51; 95%CI: 1.16-1.98) remained significant, whereas the lower
levels of extraversion were no longer significant (OR: 0.83; 95%CI: 0.64-1.08).
Type of comorbidity
To investigate whether anxiety-anxiety comorbidity is different from anxiety-depressive
comorbidity with regard to all of the aspects mentioned above, participants with anxi-
ety-anxiety comorbidity (n=105) were compared with participants with anxiety-depres-
sive comorbidity (n=292).
Participants with anxiety-anxiety comorbidity had an earlier age of onset than those
with anxiety-depressive comorbidity (19.3 years versus 22.6 years; OR: 1.33; 95%CI:
1.05-1.68) and a higher rate of chronic anxiety or depressive symptoms (53.3% versus
41.4%; OR: 0.62; CI: 0.40-0.99). The severity of anxiety and avoidance symptoms did
not significantly differ between the groups. The depressive symptoms were however
more severe among those with anxiety-depressive comorbidity (OR: 2.09; 95%CI:
1.58-2.77). Those with anxiety-depressive comorbidity had received significantly more
treatment (55.1% versus 40.0%; OR: 1.84; 95%CI: 1.17-2.90) and they also experienced
more cognitive (OR: 1.88; 95%CI: 1.44-2.46), physical (OR: 1.34; 95%CI: 1.02-1.76) and
social disability (OR: 1.46; 95%CI: 1.13-1.89).
Like the unadjusted analyses, adjusted analyses show that participants with anxi-
ety-anxiety comorbidity more often had a positive family history for anxiety or depres-
sive disorders (63.8% versus 45.9%; OR: 0.48; 95%CI: 0.30-0.76) and elevated con-
scientiousness scores (OR: 0.77; 95%CI: 0.62-0.97). The other risk-indicators did not
differ critically between both types of comorbidity.
31
Table 2
Clinical characteristics of anxiety-depressive comorbidity (n=1004).
Group (total n=1004) comparison
1 2 3 4 1 vs 2 2 vs 3 2 vs 4 3 vs 4
single anxiety anxiety-anxiety anxiety-depressive double co-
disorder comorbidity comorbidity morbidity
32
n=295 n=105 n=292 n=312 OR OR OR OR
(29.4%) (10.5%) (29.1%) (31.1%) (95% CI) (95% CI) (95% CI) (95% CI)
Clinical characteristics
Chapter 2.1
Age of onset * (years, SD) 23.0 19.3 22.6 18.0 0.73 1.33 0.88 0.66
(12.8) (11.5) (12.3) (11.3) (0.58-0.93) (1.05-1.68) (0.67-1.12) (0.59-0.77)
Chronicity (% with chronic anxie- 32.2 53.3 41.4 58.7 2.41 0.62 1.24 2.01
ty or depressive symptoms, n) (95) (56) (121) (183) (1.53-3.97) (0.40-0.99) (0.80-1.94) (1.45-2.77)
Total IDS score * (mean, SD) 19.8 23.0 29.5 37.1 1.52 2.09 4.52 2.16
(9.0) (9.6) (10.8) (11.8) (1.14-2.03) (1.58-2.77) (3.35-6.10) (1.78-2.63)
Total BAI score * (mean, SD) 12.9 16.0 16.8 23.8 1.58 1.09 2.15 1.97
(8.5) (9.3) (9.8) (11.1) (1.20-2.07) (0.85-1.42) (1.66-2.78) (1.65-2.35)
Total FQ score * (mean, SD) 27.7 33.2 31.4 44.9 1.42 0.90 1.80 2.01
(18.3) (17.5) (18.1) (20.4) (1.10-1.82) (0.70-1.15) (1.42-2.29) (1.68-2.39)
Treatment (% yes, n) 30.8 40.0 55.1 70.8 1.50 1.84 3.64 1.98
(91) (42) (161) (221) (0.94-2.37) (1.17-2.90) (2.30-5.77) (1.41-2.77)
Cognitive disability * (mean, SD) 21.0 22.5 32.9 43.5 1.11 1.88 3.28 1.74
(18.3) (17.5) (19.1) (19.2) (0.85-1.46) (1.44-2.46) (2.50-4.32) (1.46-2.08)
Physical disability * (mean, SD) 11.6 13.0 17.7 28.7 1.16 1.34 2.11 1.58
(16.6) (18.4) (20.6) (24.3) (0.86-1.55) (1.02-1.76) (1.61-2.76) (1.34-1.86)
Social disability * (mean, SD) 21.5 27.3 33.5 44.9 1.45 1.46 2.69 1.85
(16.7) (18.6) (18.5) (19.6) (1.16-1.98) (1.13-1.89) (2.07-3.50) (1.54-2.10)
IDS= Inventory of Depressive Symptomatology, BAI= Beck Anxiety Inventory, FQ= Fear Questionnaire OR; odds ratio, CI; confidence interval.
* OR per SD increase using multinomial logistic regression analyses.
Table 3
Risk indicators of anxiety-depressive comorbidity (n=1004).
Group (total n=1004) comparison

1 2 3 4 1 vs 2 2 vs 3 2 vs 4 3 vs 4
single anxiety anxiety-anxiety anxiety-depressive double
disorder comorbidity comorbidity comorbidity
n=295 n=105 n=292 n=312 OR OR OR OR
33
(29.4%) (10.5%) (29.1%) (31.1%) (95% CI) (95% CI) (95% CI) (95% CI)
Socio-demographic factors
Age (years, SD) 42.4 40.0 40.1 40.9 0.99 1.00 1.01 1.00
(13.4) (11.5) (12.5) (12.2) (0.97-1.00) (0.98-1.02) (0.99-1.02) (0.99-1.02)
Gender (% female, n) 70.8 71.4 73.6 72.1 1.03 1.12 1.03 0.93
(209) (75) 215) (225) (0.63-1.68) (0.68-1.84) (0.63-1.69) (0.95-1.33)
Education (years, SD) 12.2 12.3 11.8 10.9 1.01 0.95 0.91 0.78
(3.3) (3.0) (3.2) (3.2) (0.95-1.08) (0.89-1.02) (0.84-0.98) (0.65-0.93)
Partner (% yes, n) 67.8 64.8 67.5 65.7 0.87 1.13 1.04 0.92
(200) (68) (197) (205) (0.55-1.40) (0.71-1.80) (0.66-1.66) (0.66-1.30)
Vulnerability factors
Family history 46.4 63.8 45.9 52.6 2.07 0.48 0.74 1.31
(% yes, n) (137) (67) (134) (164) (1.30-3.30) (0.30-0.76) (0.46-1.20) (0.95-1.80)
Childhood trauma 42.4 40.0 51.0 59.3 0.91 1.56 1.50 1.09
(% yes, n) (125) (42) (149) (185) (0.58-1.43) (0.99-2.46) (0.93-2.43) (0.77-1.53)
Negative life events 0.6 0.5 0.8 0.80 0.91 1.23 1.30 1.05
(number, SD) (0.9) (0.9) (1.0) (1.1) (0.70-1.17) (0.95-1.59) (1.00-1.68) (0.90-1.22)
Neuroticism * 36.8 40.1 41.5 44.9 1.51 1.26 1.86 1.60
(total score, SD) (7.3) (7.1) (6.3) (5.9) (1.16-1.98) (0.98-1.60) (1.38-2.49) (1.30-1.97)
Extraversion * 37.3 35.4 34.5 31.5 0.83 0.87 0.67 0.70
(total score, SD) (6.4) (6.3) (6.3) (6.3) (0.63-1.08) (0.69-1.10) (0.51-0.88) (0.58-0.85)
Openness* 31.2 31.5 30.9 30.0 1.10 0.82 0.83 0.97
(total score, SD) (5.4) (5.8) (5.1) (5.5) (0.90-1.34) (0.68-1.00) (0.67-1.03) (0.86-1.11)
Agreeableness* 43.9 44.1 43.0 42.1 1.05 0.80 0.85 0.87
(total score, SD) (5.0) (5.5) (4.7) (5.7) (0.83-1.31) (0.63-1.00) (0.67-1.09) (0.73-1.03)
Conscientiousness* 37.9 37.1 35.7 37.9 0.91 0.77 1.01 0.91
(total score, SD) (5.4) (6.1) (5.8) (5.4) (0.71-1.16) (0.62-0.97) (0.78-1.32) (0.77-1.07)
OR; odds ratio, CI; confidence interval. * OR per SD increase using multinomial logistic regression analyses.
Mentioned OR are after adjustment for covariates.
2.1
Chapter 2.1
Number of comorbid disorders

To investigate whether “double” comorbidity may be distinguished from anxiety-depres-
sive comorbidity and anxiety-anxiety comorbidity, clinical factors and risk-indicators
were compared. Both anxiety-anxiety comorbidity (n=105) and anxiety-depressive co-
morbidity (n=292) were compared with the group with double comorbidity (n=312). The
differences found are presented in Tables 2 and 3.
Compared with anxiety-anxiety comorbidity, “double” comorbidity was associated
with more severe depressive (OR: 4.52; 95%CI: 3.35-6.10), anxiety (OR: 2.15; 95%CI:
1.66-2.78) and avoidance symptoms (OR: 1.80; 95%CI: 1.42-2.29). Participants with
“double” comorbidity also received more treatment (70.8% versus 40.4%; OR: 3.64;
95%CI: 2.30-5.77) and experienced more disability in all domains compared with anxi-
ety-anxiety comorbidity.
Compared with anxiety-depressive comorbidity, “double” comorbidity was associ-
ated with an earlier age of onset (18.0 versus 22.6 years; OR: 0.66; 95%CI: 0.59-0.77)
and a higher rate of chronicity (58.7% versus 41.4 %; OR: 2.01; 95%CI: 1.45-2.77).
Those with “double” comorbidity experienced more severe depressive (OR: 2.16;
95%CI: 1.78-2.63), anxiety (OR: 1.97; 95%CI: 1.65-2.35) and avoidance symptoms (OR:
2.01; 95%CI: 1.68-2.39). They also received more treatment compared with those with
anxiety-depressive comorbidity (70.8% versus 55.1%; OR: 1.98; 95%CI: 1.41-2.77) and
experienced more disability in all domains.
The unadjusted bivariate models show that, compared with those with anxiety-anxi-
ety comorbidity, participants with “double” comorbidity were less educated (10.9 versus
12.3 years) and were less likely to have a positive family history of anxiety or depres-
sive disorders (52.6% versus 63.8%). They were more likely to have experienced child-
hood trauma (59.3% versus 40.4%) and negative life events in the past year (0.8 versus
0.5 negative life events). Those with “double” comorbidity had elevated neuroticism and
conscientiousness levels and lower extraversion, openness and agreeableness levels.
(Unadjusted statistics are not shown in the Table). After adjustment for all significant
covariates in the bivariate model, the lower level of education (OR: 0.91; 95%CI: 0.84-
0.98), higher neuroticism levels (OR: 1.86; 95%CI: 1.38-2.49) and lower extraversion
scores (OR: 0.67; 95%CI: 0.51-0.88) remained significantly different between those with
“double” comorbidity and those with anxiety-anxiety comorbidity.
Comparing participants with “double” comorbidity with those with anxiety-depres-
sive comorbidity, unadjusted models show that those with “double” comorbidity were
less educated (10.9 versus 11.8 years) and more likely to have experienced childhood
trauma (59.3% versus 51.0%). They had elevated neuroticism levels and lower extra-
version and openness levels. After adjustment for all significant covariates in the bivar-
iate model, the lower level of education remained significantly different between those
with “double” comorbidity and those with anxiety-depressive comorbidity (OR: 0.78;
95%CI: 0.65-0.93), as did the elevated neuroticism levels (OR: 1.60; 95%CI: 1.30-1.97)
and decreased extraversion levels (OR: 0.79; 95%CI: 0.58-0.85).
DISCUSSION
Clinical characteristics of comorbidity
In a well-defined subsample of the NESDA study of 1004 participants with at least one
34
anxiety disorder, the clinical relevance of comorbidity was studied. We may conclude
that both the type of comorbidity and the number of comorbid disorders are clinically
important to diagnose.
Anxiety-anxiety comorbidity was associated with more severe clinical characteris-
tics compared with a single anxiety disorder. Patients with anxiety-anxiety comorbidity
experienced more severe symptoms, more chronic symptoms and more social impair-
ment. It should be noted that in anxiety-anxiety comorbidity, not only anxiety symp-
toms were more severe, depressive symptoms were also more severe. These findings
confirm an earlier finding of an association between anxiety-anxiety comorbidity and
severity of symptoms.13 The finding that anxiety-anxiety comorbidity is associated with 2.1
more social impairment contradicts earlier research that could not find an association
between the number of anxiety disorders and the quality of life.14-16
Anxiety-anxiety comorbidity was associated with critically different clinical factors
compared with anxiety-depression comorbidity: The course of anxiety-anxiety comor-
bidity was worse than the course of anxiety-depressive comorbidity, with an earlier age
of onset and more chronicity. anxiety-depressive comorbidity, however, is associated
with more treatment and higher levels of disability. These results suggest that the
type of comorbidity is important to diagnose, because it differentiates anxiety-anxiety
comorbidity from anxiety-depressive comorbidity: the first type is more chronic, the
second type is more invalidating. “Double” comorbidity in anxiety disorders appeared
to lead to more severe symptoms, more treatment and more day-to-day impairment
compared with anxiety-anxiety comorbidity or anxiety-depressive comorbidity. This
indicates that the number of diagnoses present is related to severity, use of health care
and negative consequences.
Sociodemographics and vulnerability factors of comorbidity

After adjusting for covariates, this study found that neuroticism is the most consistent
risk-indicator associated with comorbidity in anxiety disorders. Neuroticism was lowest
in patients with a single anxiety disorder, higher in patients with anxiety-anxiety comor-
bidity and anxiety-depressive comorbidity and highest in those with double comorbidi-
ty. This finding may not be specific for anxiety disorders but may reflect general vulner-
ability given that more psychiatric disorders are associated with high neuroticism.35-37
Ormel and colleagues38 argue that the association of neuroticism and psychopathology
is tautological, since neuroticism reflects the presence of subthreshold anxiety and de-
pressed feelings over a period of time, which obviously predict the development of an
anxiety or depressive disorder.
Limitations and strengths

Several strengths and limitations should be taken into account. Strengths include the
large sample size, the well-implemented design of the NESDA study and the use of
the CIDI as a diagnostic instrument safeguarding statistical power and the reliability
of results. Limitations include the cross-sectional design. A prospective design would
be more appropriate to study the outcome. Another potential limitation of this study is
Post Traumatic Stress Disorder, Obsessive-Compulsive Disorder and Specific Phobia
were not assessed in the NESDA study. Moreover, the single anxiety disorders that
35
Chapter 2.1
were assessed were collapsed into one group to achieve sufficient statistical power. If
these disorders had had different sociodemographics, vulnerability factors or clinical
characteristics, analyzing anxiety disorders as one group could have resulted in a bias.
Therefore, we have compared the specific anxiety disorders with each other (data not
shown). All anxiety disorders appeared to be similar. Based on this, we feel it is justi-
fied to examine anxiety disorders as one group. Finally, in this study no distinction was
made between whether depression or anxiety disorder preceded the current anxiety
disorder, thus hampering establishment of whether or not comorbid depression might
be secondary to the anxiety disorder. However, a recent study39 found no differences in
clinical characteristics between comorbidity with preceding depression and comorbidi-
ty with preceding anxiety.
Conclusion
This study adds to current knowledge on comorbidity in anxiety disorders. Its results
clearly showed that it is important to diagnose comorbidity among anxiety disorders as
it is associated with higher severity and more chronicity compared with single anxiety
disorders. In addition, it is also important to distinguish the type and number of comor-
bid disorders: Whereas anxiety-anxiety comorbidity has an earlier age of onset and
a more chronic course, anxiety-depressive comorbidity leads to more treatment and
impaired functioning. Furthermore, “double” comorbidity leads to even more severity,
chronicity and impairment compared with both anxiety-anxiety and anxiety-depressive
comorbidity. These findings indicate that in clinical practice all comorbid disorders
present in patients with an anxiety disorder should be diagnosed. Given the more
severe symptomatology, unfavorable course and impaired functioning, patients with
an anxiety disorder and comorbid anxiety and depressive symptoms should receive
prompt and adequate treatment.
Acknowledgements
The infrastructure for the NESDA study is supported by participating universities and
mental health care organizations (VU University Medical Center, GGZinGeest, Arkin,
Leiden University Medical Center, GGZ Rivierduinen, University Medical Center Gron-
ingen, Lentis, GGZ Friesland, GGZ Drenthe, IQ Healthcare, Netherlands Institute for
Health Services Research (NIVEL) and Netherlands Institute of Mental Health and Ad-
diction (Trimbos).
36
Reference List
2001;110:585-599.
2. Brown TA, Barlow DH. Comorbidity among anxiety and mood disorders: Implica-
tions for treatment and DSM-IV. J Consult and Clin Psychol 1995;63:408-441.
4. de Graaf R, Bijl RV, Smit F et al. Risk factors for 12-month comorbidity of mood, 2.1
anxiety, and substance use disorders: findings from the Netherlands Mental Health
Survey and Incidence Study. Am J Psychiatry 2002;159:620-629.
5. Alonso J, Angermeyer, MC, Bernert S et al. 12-Month comorbidity patterns and as-
sociated factors in Europe: results from the European Study of the Epidemiology of
Mental Disorders (ESEMeD) project. Acta Psychiatr Scand Suppl 2004;420:28-37.
1991;84:446-452.
Arch Gen Psych 2003;60:993-1000.
2005;87:43-55.
9. Kessler RC, McGonagle KA, Zhao S et al. Lifetime and 12-month prevalence of
DSM-III-R psychiatric disorders in the United States. Results from the National Co-
morbidity Survey. Arch Gen Psychiatry 1994;51:8-19.
chol Med 1997;27:861-873.
11. Van Balkom AJ, Boeijen CA, van Boeke AJ et al. Comorbid depression, but not
comorbid anxiety disorders, predicts poor outcome in anxiety disorders. Depress
Anxiety 2008;25:408-415.
12. Chambers JA, Power KG, Durham RC. The relationship between trait vulnerability
and anxiety and depressive diagnoses at long-term follow-up of Generalized Anxi-
ety Disorder. J Anxiety Disord 2004;18:587-607.
13. Cougle JR, Timpano KR, Sachs-Ericsson N et al. Examining the unique relation-
ship between anxiety disorders and childhood physical and sexual abuse in the
National Comorbidity Survey-Replication. Psychiatry Res 2010;177:150-155.
14. Grant BF, Hasin DS, Stinson FS et al. Prevalence, correlates, co-morbidity, and
comparative disability of DSM-IV generalized anxiety disorder in the USA: results
from the National Epidemiologic Survey on Alcohol and Related Conditions. Psy-
chol Med 2005;35:1747-1759.
15. Norberg MM, Diefenbach GJ, Tolin DF. Quality of life and anxiety and depressive
disorder comorbidity. J Anxiety Disord 2008;22:1516-1522.
16. Barrera TL, Norton PJ. Quality of life impairment in generalized anxiety disorder,
social phobia and panic disorder. J Anxiety Disord 2009;23:1086-1080.
37
Chapter 2.1
17. Penninx BW, Beekman AT, Smit JH et al. The Netherlands Study of Depression
Res 2008;17:121-140.
18. World Health Organization. Composite International Diagnostic Interview (CIDI):
version 2.1. Geneva, Switzerland: 1997.
19. Wittchen HU. Reliability and validity studies of the WHO-Composite International
Diagnostic Interview (CIDI): a critical review. J Psychiatr Research 1994;28:57-84.
20. Fyer AJ, Weissman MM. Genetic linkage study of panic: clinical methodology and
description of pedigrees. Am J Med Genet 1999;88:173–181.
21. Brugha T, Bebbington P, Tennant C et al. The List of Threatening Experiences: a
subset of 12 life event categories with considerable long-term contextual threat.
Psychol Med 1985;15:189-194.
22. Brugha T, Cragg D. The list of threatening experiences: the reliability and validity of
a brief live events questionnaire. Acta Psychiatr. Scand 1990;82:77–81.
23. Costa PT, McCrae RR. Domains and facets: hierarchical personality assessment
using the revised NEO personality inventory. J Pers Assess 1995;64:21-50.
24. Costa PT, McCrae RR. Revised NEO. Personality Inventory(NEO-PR-I) and the
Five Factor Inventory (NEO-FFI): Professional Manual. Psychological Assessment
Resources, Odessa, FL: 1992.
25. Lyketsos CG, Nestadt G, Cwi J et al. The life chart interview: a standarized
method to describe the course of psychopathology. Int J Methods Psychiatr Res
1994;4:143-155.
26. Warshaw MG, Keller MB, Stout RL. Reliability and validity of thelongitudinal inter-
val follow-up evaluation for assessing outcome of anxiety disorders. J. Psychiatr.
Res 1994;28:531–545.
27. Hakkaart-van Roijen L. Trimbos/iMTA questionnaire for costs associated with psy-
chiatric illness (TIC-P). Rotterdam: Institute for Medical Technology Assessment:
2002.
28. Beck AT, Epstein N, Brown G et al. An inventory for measuring clinical anxiety:
psychometric properties. J Consult and Clin Psychol 1988;56:893-897.
29. Fydrich T, Dowdall D, Chambless DL. Reliability and Validity of the Beck Anxiety
Inventory. J Anxiety Disord 1992;6:55–61.
30. Marks IM, Matthews AM. Brief standard self-rating for phobic patients. Behav Res
Ther 1979;17:263-267.
31. Van Zuuren FJ. The fear questionnaire. Some data on validity, reliability and lay-
out. British J Psychiatry 1988;153:659-662.
32. Rush AJ, Gullion CM, Basco MR et al. The Inventory of Depressive Symptomatolo-
gy (IDS): psychometric properties. Psycholog Med 1996;26:477-486.
33. Trivedi MH, Rush AJ, Ibrahim HM et al. The Inventory of Depressive Symptomatol-
ogy, Clinician Rating (IDS-C) and Self-Report (IDS-SR), and the Quick Inventory
of Depressive Symptomatology, Clinician Rating (QIDS-C) and Self-Report (QIDS-
SR) in public sector patients with mood disorders: a psychometric evaluation. Psy-
cholog Med 2004;34:73-82.
34. Chwastiak LA, Von Kurff M. Disability in depression and back pain: evaluation of
the World Health Organization Disability Assessment Schedule (WHO DAS II) in a
38
primary care setting. J Clin Epidemiol 2003;56:507-514.

35. Enns MW, Cox BJ. Personality dimensions and depression: review and commen-
tary. Can J Psychiatry 1997;42:274-284.
36. Van Os J, Jones PB. Neuroticism as a risk factor for schizophrenia. Psychol Med
2001;31:1129–1134.
37. Cox BJ, MacPherson PS, Enns MW et al. Neuroticism and self-criticism associat-
ed with posttraumatic stress disorder in a nationally representative sample. Behav
Res Ther 2004;42:105–114.
38. Ormel J, Rosmalen J, Farmer A. Neuroticism: a non-informative marker of vulnera-
bility to psychopathology. Soc Psychiatry Psychiatr Epidemiol 2004;39:906-912. 2.1
39. Lamers F, van Oppen P, Comijs HC et al. Comorbidity patterns of anxiety and de-
39
40
Chapter 2.2
Clinical relevance of comorbidity in Obsessive Compulsive

Disorder: The Netherlands OCD Association Study.
Mieke Klein Hofmeijer-Sevink, Patricia van Oppen, Harold J. van Megen, Neeltje
M. Batelaan, Danielle C. Cath, Nic J.A. van der Wee, Marcel A. van den Hout, An-
ton J. van Balkom.
Journal of Affective Disorders 2013; 150:847-854.
Author contributions: Authors P van Oppen, H van Megen, D Cath, N van der Wee and A van Balkom con-
tributed to the design of the study. The study is being coordinated by P van Oppen. The present manuscript
was drafted by H van Megen, N Batelaan, D Cath, N van der Wee, A van Balkom, M van der Hout and P van
Oppen. Author Klein Hofmeijer-Sevink managed the literature searches, the analyses and she wrote the first
draft of the manuscript. All authors contributed to critical revision of the manuscript for important intellectual
content. All authors read and approved the final manuscript.
41
Chapter 2.2
Abstract
Background: This study describes lifetime and current rates of comorbidity, its onset
and its consequences in a large clinical sample of patients with obsessive compulsive
disorder (OCD). A wide range of risk factors and clinical characteristics were also ex-
amined to determine whether pure OCD is different from OCD with current comorbidity.
Finally, the temporal sequencing of the disorders was examined.
Method: Data were obtained from the Netherlands Obsessive Compulsive Disorder
Association (NOCDA) study. A sample of 382 participants with current OCD (during the
past month) was evaluated.
Results: Current comorbidity occurred in 55% of patients with OCD, while 78% suf-
fered from lifetime comorbidity. Comorbidity is associated with more severe OCD, anx-
iety and depressive symptoms and more negative consequences on daily life. Multiple
comorbid disorders often precede OCD and influence both its course and severity.
Childhood trauma and neuroticism are vulnerability factors for the development of mul-
tiple comorbid disorders in OCD.
Limitations: It should be noted that causal inferences about the association between
risk factors and OCD are precluded since our results were based on cross-sectional
data.
Conclusion: (Multiple) comorbidity in OCD is clinically relevant since it is associated
with a specific pattern of vulnerability, with greater chronicity, with more severe OCD
and more negative consequences on daily life. This indicates that the diagnosis and
treatment of all comorbid disorders is clinically relevant, and clinicians should be es-
pecially aware of multiple disorders in cases of childhood trauma and high levels of
neuroticism. Primary OCD has a different developmental and comorbidity pattern com-
pared to secondary OCD.
42
Comorbiditiy in OCD
INTRODUCTION
Obsessive Compulsive Disorder (OCD) is a disabling and often chronic disorder1 with
a lifetime prevalence of around 2%.2 Comorbidity in OCD is the rule rather than the
exception.3 Lifetime comorbidity rates have been estimated at around 90%4 and one-
month comorbidity rates at around 40%. 4,5
While the clinical relevance of comorbidity in other anxiety disorders has been
widely investigated, research on the clinical relevance of comorbidity in OCD is rela-
tively scarce. The available data are often not presented in studies that were designed
to investigate comorbidity, but in studies with a different primary focus (for example
treatment outcomes) that include comorbidity as a clinical,6,7 in studies with relatively
small samples (n<74)8,9 or in studies that focus either on clinical characteristics6 or vul-
nerability factors,10 but not on all these variables simultaneously. To our knowledge, two
previous studies have examined comorbidity in a large, broad-ranging OCD sample.5,11
2.2
Furthermore, the results from these previous studies on comorbidity in clinical OCD
samples are inconsistent. There is some evidence that compared to ‘pure’ OCD, comor-
bidity in OCD is associated with poorer global functioning,5 poorer treatment outcomes,6
increased chronicity,6 increased OCD symptom severity11 and with the specific subtype
of hoarding in OCD.7 However, other studies do not support these findings.5,8-10,12
Another understudied area in OCD comorbidity research is temporal sequencing.
In temporal sequencing, the focus is on whether OCD precedes the comorbid dis-
order or occurs after the onset of the comorbid disorder. One study found that OCD
precedes the comorbid disorder in 85% of the cases.13 More recent data from the Na-
tional Comorbidity Survey replication suggested that a comorbid depressive disorder
often occurs after the onset of OCD, while a comorbid anxiety disorder often precedes
OCD.14
To sum up, there is a lack of large-scale cohort studies that systematically examine
the clinical relevance of comorbidity in OCD. Moreover, the smaller studies that have
investigated this theme have been somewhat inconsistent. This study aims to add to
current knowledge by exploring the many clinical factors and risk indicators associated
with comorbidity in OCD in a large sample.
Using data from the Netherlands Obsessive Compulsive Disorder Association
(NOCDA) study, this study aims to describe the lifetime and current occurrence of co-
morbidity, its onset and its consequences in a large clinical sample of OCD patients.
We also aimed to examine a wide range of risk factors as well as clinical characteristics
to determine whether pure OCD differs from OCD with current comorbidity. Finally, the
temporal sequence of OCD and comorbid disorders was examined.
Our hypotheses are that comorbidity is associated with a larger number of risk
indicators (socio-demographic, vulnerability and personality factors) and more severe
clinical factors, compared to pure OCD.
METHOD
Sample
Data were obtained from the Netherlands Obsessive Compulsive Disorder Association
(NOCDA) study. The detailed methodology of the NOCDA has been described else-
where.15 In brief, the NOCDA study entails an ongoing naturalistic longitudinal cohort
43
Chapter 2.2
study conducted among 419 adult patients with a lifetime OCD, of whom 382 met the
criteria for having OCD during the past month. Because the NOCDA study is a natural-
istic study, exclusion criteria were limited to an inadequate understanding of the Dutch
language for the purposes of completing the interviews and self-reporting question-
naires. All the patients diagnosed with OCD who were referred to one of the participat-
ing mental health care centers were asked for permission to be contacted for research
purposes during the inclusion procedure. All those OCD patients who consented were
contacted and invited to participate in the study, irrespective of the stage of the disor-
der, the OCD subtype, the presence of comorbidity and the stage of chronicity.
Multiple centers in the Netherlands participated and patients were followed for a
period of six years with one wave of data collection carried out one year after baseline,
and subsequent biannual measurements. These procedures were approved by the
ethics committee of all participating centers and written informed consent was obtained
from all participants. For the present study we used baseline data, which included an
interview, a medical assessment and self-administered questionnaires.
Diagnostic measures
To diagnose DSM-IV axis I disorders, the Structured Clinical Interview for DSM-IV-TR
(SCID-I/P) was used.16 The SCID-I/P is a semi-structured interview for diagnosing men-
tal disorders. Both one-month and lifetime diagnoses were assessed. The Dutch ver-
sion of the SCID-I/P has good inter-rater reliability rates.17 To ensure the quality of the
data, all the interviewers were trained and supervised to conduct the SCID-I/P.
Temporal sequencing was retrospectively determined using data on the age of on-
set as assessed by the SCID-I/P. A variable with three categories was constructed: 1)
OCD was the primary diagnosis when the onset of OCD preceded the onset of the co-
morbid diagnosis, 2) OCD and the comorbid disorder had a simultaneous onset when
both had the same age of onset and 3) the comorbid diagnosis was the primary disor-
der when the onset of the comorbid diagnosis preceded the onset of OCD.
Correlates of comorbidity
Clinical characteristics: The severity of the OC symptoms was assessed using the
Padua Inventory-Revised (PI-R)18 and the Yale-Brown Obsessive-Compulsive Scale
(Y-BOCS, a severity scale) and Y- BOCS-symptom checklist (Y-BOCS-SC).19 The PI-R
was derived from the Padua Inventory20 and was used to assess the overall severity of
OCD symptoms and the presence and severity of specific OCD sub scales. The reli-
ability and validity of this instrument are well-established.18 The Y-BOCS severity scale
is a 10-item rater-administered measure of the current severity of obsessions and com-
pulsions. This scale is a reliable and valid instrument for assessing the severity of OCD
symptoms.21 A minimally adapted version of the Y-BOCS-symptoms checklist was used
to assess specific obsessions and compulsions and to categorize all OCD symptoms
into one or more of four separate symptom dimensions: ‘Aggressive, sexual, religious
and somatic obsessions, checking compulsions’ (20 items), ‘Symmetry obsessions, re-
peating, counting and ordering compulsions’ (10 items), ‘Contamination obsessions and
cleaning compulsions’ (9 items) and ‘hoarding obsessions and compulsions’ (2 items).22
Both the presence of the dimensions (yes/no) and their severity were calculated.
44
Comorbiditiy in OCD
The version of the Y-BOCS-SC used in our study is an adapted self-report version
of the original interview version.19 In our adapted Y-BOCS-SC we used a self-report
version and thirteen items were excluded. Both changes were according to the version
of Bear.23 Furthermore six items from different dimensions were left out from the orig-
inal version. Factor analysis of our adapted version of the original Y-BOCS-SC led to
the same symptom dimensions as the original list.24
The age of onset was determined with the SCID-I/P. Following recent literature, on-
set was considered ‘early’ when the OCD started before the age of 20 years.25 Patients
were asked if they had used antidepressants for OCD symptoms during the past two
weeks. Treatment delay was defined as the time elapsed between the age of onset of
OCD and the first treatment (both retrospectively estimated by the patient). The chro-
nicity of OCD symptoms was based on the Life Chart Interview in which the presence
of OCD symptoms was retrospectively assessed. Chronic OCD was defined as the
2.2
continuous presence of at least moderately severe OCD symptoms for at least two
years prior to the baseline measurement.
The severity of the depressive and anxiety symptoms were assessed with the Beck
Depression Inventory (BDI)26 and the Beck Anxiety Inventory (BAI).27 Both are 21-item
self-report lists that are widely used and have proven highly valid and reliable.28,29
Quality of life was assessed with the EuroQol.30 This questionnaire assesses the
consequences of disease on five aspects of quality of life (mobility, self-care, daily ac-
tivities, pain/discomfort and anxious/depressive feelings). This instrument has shown
itself to be suitable and reliable in the general population.31 Patients also had to rate
their health-related quality of life using a visual analogue scale (ranging from 0-100).
Socio-demographic factors included age (in years), gender (male/female), partner
status (absent/present) and education (in years).
Vulnerability factors included childhood trauma, family history of OCD, negative life
events and personality traits. Childhood trauma was assessed using parts of the Struc-
tured Trauma Interview32 to address childhood physical abuse and sexual abuse. Child-
hood trauma was considered present in case of reporting physical- and/or sexual abuse
before the age of 16. An assessment of family history of OCD according to the partici-
pants was considered positive if one or more first-degree family members suffered from
OCD. Negative life events in the past year were based on a list of twelve negative life
events, counting the number of negative life events in the past year. Personality traits
were measured with the 100-item Five-Factor Personality Inventory (FFPI)33 which uses
five dimensions for personality: extraversion, agreeableness, conscientiousness, neu-
roticism and autonomy. The FFPI has been proven reliable and valid.34
Statistical analysis
All statistical analyses were conducted with the SPSS statistical package version 18.0.
First, within the sample of patients with current OCD, the comorbidity rates of one
month and lifetime axis I diagnoses were determined.
Second, patients with pure OCD (without current comorbidity) were compared with
patients with OCD with current comorbidity for risk indicators and clinical character-
istics. Bivariate logistic regression analyses were conducted to examine the clinical
factors of comorbidity. After Bonferroni’s correction, characteristics with a p<0.002
45
Chapter 2.2
were considered significant. Subsequently, bivariate and multivariate logistic regres-

sion analyses were conducted to examine risk indicators. After Bonferroni’s correction,
characteristics with a p<0.004 were considered significant. In the multivariate model,
we adjusted for all covariates that were significant in the bivariate analyses.
Finally, a sub-analysis was performed to examine the sample on the number of
comorbid disorders. To achieve sufficient statistical power, participants with two or
more comorbid disorders were combined into one group. Pure OCD was taken as the
reference group and then compared to OCD with one comorbid disorder and to OCD
with multiple comorbid disorders. Subsequently, OCD with one comorbid disorder was
taken as the reference group and compared to OCD with multiple comorbid disorders.
Both bivariate and multivariate multinomial logistic regression analyses were conduct-
ed. In the multivariate model, we adjusted for all risk indicators that were significant in
the bivariate analyses.
RESULTS
Comorbidity rates
The sample of 382 patients (57% female) with a current OCD had a mean age of 36.4
years (SD:10.9 years). The majority of the sample were living with a partner (59%), they
had undergone a mean of 12.6 years of education and had a mean Y-BOCS score of
21.2 (SD:7.0), indicating moderately severe OCD. The majority of the sample (62%)
were currently using antidepressant medication for OCD symptoms.
Of all current OCD patients (n=382), 45% (n=172) had a pure OCD while 55%
(n=210) suffered from at least one current comorbid disorder. Of those suffering from
comorbidity, 119 patients had one comorbid disorder while 91 patients suffered from
multiple comorbid disorders (Table 1). The mean number of comorbid disorders (in the
multiple comorbidity group) was 2.6 (SD: 0.98). Anxiety- and mood disorders were the
most prevalent comorbid disorders. A current comorbid anxiety disorder was present in
37% (n=140) of the patients and a current comorbid mood disorder was present in 25%
(n=95). Other comorbid axis I diagnoses such as psychotic disorders, eating disorders
and substance use disorders were less common (<10%).
The majority of the OCD patients (78%, n= 297) had a lifetime history of any comor-
bid disorder: 46% (n=176) with a lifetime history of anxiety disorders and 64% (n=244)
with a history of mood disorders.
Comparison of pure OCD to OCD with current comorbidity

To investigate whether pure OCD differs from OCD with current comorbidity, patients
with pure OCD (n=172) were compared to patients with OCD with current comorbidity
(n=210, Table 2). The severity of OCD symptoms measured with the Y-BOCS did not
differ significantly between pure OCD and OCD with current comorbidity. Patients with
OCD with current comorbidity reported significantly more severe checking symptoms.
The total PI-R score was significantly higher among patients with OCD and comorbidi-
ty, as were the PI-R scores on the subscales ‘precision’, ‘rumination’ and ‘impulses’.
Patients with OCD with current comorbidity were comparable with patients with
pure OCD in terms of the age of onset of OCD, the percentage with early onset, treat-
ment delay and chronicity. Comorbidity was not associated with greater use of antide-
46
Comorbiditiy in OCD
Table 1
Current and lifetime comorbidity rates among one month OCD patients (n=382).
Rates*
Current Lifetime
Diagnosis % (n) % (n)
Pure OCD 45.0 (172) 22.3 (85)
OCD + any Anxiety Disorder 36.6 (140) 46.1 (176)
+ SP 19.1 (73) 24.1 (92)
+ PD 9.2 (35) 19.1 (73)
+ GAD 9.2 (35) 9.2 (35) 2.2

+ Specific Phobia 8.2 (32) 10.5 (40)
+ PTSD 3.1 (12) 4.7 (18)
+ AGO 1.3 (5) 2.6 (10)
+ Other Anxiety Disorder 0.5 (2) 0.5 (2)
OCD + any Mood Disorder 24.9 (95) 63.9 (244)
+ MDD 18.3 (70) 56.5 (216)
+ DYS 5.5 (21) 5.5 (21)
+ BIP 1.0 (4) 3.4 (13)
OCD + Somatoform Disorder 5.8 (22) 5.8 (22)
OCD + any Substance Use Disorder 5.2 (20) 13.6 (52)
+ Substance Dependence 4.2 (16) 10.7 (41)
+ Substance Abuse 1.0 (4) 3.7 (14)
OCD+ Eating Disorder 5.0 (19) 10.7 (41)
OCD + Psychotic Disorder 2.6 (10) 4.7 (18)
* The sum is more than 100% (n=382) because some patients had more than 1 disorder.
OCD: obsessive-compulsive disorder, PD: panic disorder (with or without agoraphobia), GAD: gene-
ralized anxiety disorder, SP: social phobia, PTSD: post-traumatic stress disorder, AGO: Agoraphobia
(without panic disorder), MDD: major depressive disorder, DYS: dysthymic disorder, BIP: bipolar
disorder.
pressants. As might have been expected, patients with OCD with current comorbidity
had significantly more severe anxiety and depressive symptoms. They also rated their
health-related quality of life significantly lower and experienced significantly more
negative consequences regarding usual activities, pain or discomfort and anxious or
depressive feelings.
47
Table 2
Clinical characteristics of patients with pure OCD versus comorbidity (n=382).
Current diagnosis (n=382) Pure OCD OCD with comorbidity OR (CI 95%)
(45%, n=172) (55%, n=210)
Clinical factors
48
Age of onset OCD (years; SD) 17.9 8.5 18.1 (9.3) 1.00 (0.98-1.03)
Early onset (% yes) 69.2 64.3 0.80 (0.52-1.23)
Use of antidepressant (% yes) 57.6 64.8 1.36 (0.90-2.05)
Chapter 2.2
Treatment delay (years, SD) 9.0 (8.0) 10.4 (9.5) 1.02 (1.00-1.04)
Chronicity (% yes) * 59.9 63.8 1.18 (0.78-1.79)
BAI score (mean, SD) * 12.9 (8.5) 22.5 (12.2) 2.92 (2.19-3.90)^
BDI score (mean, SD) * 11.6 (7.1) 19.7 (10.1) 2.98 (2.23-3.98)^
Y BOCS score (mean, SD) * 20.5 (6.4) 21.8 (7.5) 1.23 (0.97-1.55)
PI-R total score (mean, SD) * 54.2 (22.1) 68.4 (28.4) 1.83 (1.44-2.31)^
Dimension Y BOCS**
Check: % yes (n) 90.1 (155) 95.7 (201) 2.45 (1.06-5.64)
Severity (range 1-20, SD)* 4.3 (2.6) 5.6 (3.6) 1.56 (1.25-1.93)^
Sym: % yes (n) 70.9 (122) 73.3 (154) 1.13 (0.72-1.77)
Severity (range 1-10, SD)* 3.0 (2.2) 4.0 (2.7) 1.17 (0.97-1.41)
Cont: % yes (n) 63.4 (109) 67.6 (142) 1.21 (0.79-1.84)
Hoar: % yes (n) 20.3 (35) 20.5 (43) 1.01 (0.61-1.66)
PI-R score (score, SD)
Washing* 10.9 10.2 12.5 11.8 1.17 (0.94-1.47)
Checking* 13.2 7.0 14.9 7.6 1.55 (1.05-2.29)
Precision* 5.9 5.4 7.8 6.3 1.45 (1.13-1.86)
Rumination* 20.4 7.5 25.3 8.4 5.69 (3.04-10.63)^
Impulses* 3.7 3.9 8.0 7.1 2.36 (1.79-3.11)^
Quality of life (% yes)
Problems with mobility 5.2 11.9 2.45 (1.11-5.40)
Problems with self-care 7.0 9.5 1.40 (0.67-2.97)
Problems with usual activities 47.1 63.3 1.94 (1.29-2.93)^
Pain or discomfort 30.8 51.0 2.33 (1.53-3.56)^
49
Anxious/depressive feelings 58.1 81.9 3.26 (2.05-5.18)^
Health related quality of life (SD) 70.6 (15.0) 61.4 (19.8) 0.58 (0.46-0.72)^
Risk indicators
Socio-demographics
Age (years, SD) 35.7 (10.6) 36.9 (11.2) 1.01 (0.99-1.03)
Gender (% female) 54.1 58.6 1.20 (0.80-1.80)
Comorbiditiy in OCD
Partner (% yes) 59.9 59.0 0.97 (0.64-1.46)

Education(years;SD) * 12.8 (3.1) 12.4 (3.3) 0.90 (0.73-1.10)
Childhood trauma (% yes) 8.7 17.6 2.24 (1.18-4.24)
Family history OCD (% yes) 39.5 44.3 1.22 (0.81-1.83)
Number of negative life events (SD) 1.5 (1.4) 1.8 (1.5) 1.15 (1.00-1.33)
Personality factors (score,SD)*
Extraversion 67.6 (13.0) 59.3 (14.4) 0.53 (0.42-0.66)^
Agreeableness 78.0 (9.2) 79.0 (10.7) 1.11 (0.90-1.35)
Conscientiousness 71.1 (10.6) 70.5 (11.9) 0.95 (0.78-1.16)
Neuroticism 37.5 (10.8) 47.8 (11.1) 2.98 (2.20-3.80)^
Autonomy 67.1 (10.3) 61.0 (11.1) 0.54 (0.43-0.69)^
OR;odds ratio, CI;confidence interval, SD=standard deviation, BAI=beck anxiety inventory, BDI=beck depression inventory, Y BOCS=Yale Brown Obsessi-
ve-Compulsive Scale, PI-R:Padua Inventory-Revised.
^ p = significant (due to Bonferroni correction significance level was set on 0.002 for clinical factors and on 0.004 for risk indicators)
* OR per SD increase using bivariate logistic regression analyses (Mentioned odds ratios and confidence intervals are unadjusted. Significant findings of
adjusted data are mentioned in the article.)
** Dimensions: Check: Aggressive, sexual, religious and somatic obsessions, checking compulsions; Sym: symmetry obsessions, repeating, counting and
ordering compulsions; Cont: contamination obsessions and cleaning compulsions; Hoar: hoarding obsessions and compulsions;
2.2
Table 3
Pure OCD versus OCD with one comorbid disorder and versus OCD with multiple comorbid disorders (n=382)
Current diagnosis (n=382) Group 1. Group 2. Group 3. OR (CI 95%)

Pure OCD OCD with one OCD with
comorbid disorder multiple comorbid
disorders
(45%, n=172) (31.2%, n=119) (23.8%, n=91) 1 vs 2 1 vs 3 2 vs 3
50
Clinical factors
Age of onset OCD (years; SD) 17.9 (8.5) 18.3 (9.4) 18.0 (9.1) 1.01 (0.98-1.03) 1.00 (0.97-1.03) 1.00 (0.97-1.03)
Early onset (% yes) 69.2 65.9 63.0 0.76 (0.64-1.24) 0.86 (0.50-1.48) 1.14 (0.64-2.01)
Chapter 2.2
use of antidepressant (% yes) 57.6 59.7 71.4 1.09 (0.68-1.75) 1.84 (1.07-3.18) 1.69 (0.94-3.03)
Treatment delay (years, SD) 9.0 (8.0) 9.8 (9.1) 11.1 (9.7) 1.01 (0.99-1.04) 1.03 (0.99-1.06) 1.01 (0.98-1.40)
Chronicity (% yes) * 59.9 53.8 67.9 0.78 (0.49-1.25) 2.23 (1.26-3.97)^ 2.87 (1.56-5.25)^
BAI score (mean, SD) * 12.9 (8.5) 18.6 (10.9) 27.5 (12.1) 2.20 (1.60-3.01)^ 4.57 (3.20-6.52)^ 2.18 (1.60-2.97)^
BDI score (mean, SD) * 11.6 (7.1) 17.1 (9.0) 23.1 (10.4) 2.36 (1.73-3.22)^ 4.33 (3.04-6.15)^ 1.88 (1.39-2.55)^
YBOCS score (mean, SD) * 20.5 (6.4) 20.7 (7.4) 23.2 (7.4) 1.03 (0.78-1.35) 1.56 (1.16-2.11)^ 1.45 (1.07-1.97)
PI-R total score (mean, SD) * 54.2 (22.1) 61.9 (26.2) 77.0 (28.9) 1.44 (1.10-1.88) 2.55 (1.90-3.43)^ 1.73 (1.29-2.30)^
Quality of life(% yes)
Problems with mobility 5.2 10.9 13.2 2.22 (0.92-5.38) 2.75 (1.11-6.80)^ 1.24 (0.54-2.86)
Problems with self-care 7.0 6.7 13.2 0.96 (0.38-2.43) 2.03 (0.87-4.71) 2.11 (0.82-5.40)
Problems with usual activities 47.1 55.5 73.6 1.40 (0.88-2.23) 3.14 (1.80-5.46)^ 2.24 (1.24-4.04)
Pain or discomfort 30.8 41.2 63.7 1.57 (0.97-2.56) 3.95 (2.31-6.75)^ 2.51 (1.43-4.41)
Anxious/depressive feelings 58.1 77.3 87.9 2.45 (1.45-4.15)^ 5.24 (2.60-10.54)^ 2.13 (1.00-4.58)
Health related quality of life(SD) 70.6 (15.0) 64.7 (17.9) 57.1 (51.3) 0.68 (0.52-0.88)^ 0.47 (0.35-0.61)^ 0.69 (0.53-0.90)
Risk indicators
Socio-demographics
Age (years, SD) 35.7 (10.6) 36.3 (10.9) 37.6 (11.5) 1.01 (0.98-1.03) 1.02 0.99-1.04) 1.01 (0.99-1.04)
Gender (% female) 54.1 55.5 62.6 1.06 (0.66-1.69) 1.42 (0.85-2.40) 1.35 (0.77-2.35)
Partner (% yes) 59.9 60.5 57.1 1.03 (0.64-1.65) 0.89 (0.53-1.50) 0.87 (0.50-1.52)
Education(years;SD) * 12.8 (3.1) 12.6 (3.4) 12.2 (3.1) 0.99 (0.92-1.06) 0.95 (0.88-1.03) 0.96 (0.88-1.04)
Childhood trauma (% yes) 8.7 13.4 23.1 1.63 (0.77-3.43) 3.14 (1.53-6.45)^ 1.93 (0.94-3.96)
Family history OCD (% yes) 39.5 42.0 47.3 1.11 (0.69-1.78) 1.37 (0.82-2.29) 0.81 (0.47-1.40)
Number of negative life events 1.5 (1.4) 1.8 (1.6) 1.7 (1.4) 1.18 (1.01-1.38) 1.13 (0.95-1.35) 0.96 (0.80-1.14)
(SD)
Personality factors (score, SD)*
Extraversion 67.6 (13.0) 61.0 (13.5) 57.0 (15.3) 0.60 (0.46-0.77)^ 0.45 (0.33-0.59)^ 0.76 (0.57-1.00)
51
Agreeableness 78.0 (9.2) 79.2 (10.8) 78.9 (10.5) 1.12 (0.89-1.42) 1.09 (0.84-1.40) 0.98 (0.76-1.26)
Conscientiousness 71.1 (10.6) 70.3 (11.8) 71.1 (10.6) 0.93 (0.74-1.18) 0.97 (0.75-1.25) 1.04 (0.80-1.35)
Neuroticism 37.5 (10.8) 46.7 (11.0) 49.3 (11.1) 2.59 (1.93-3.49)^ 3.40 (2.44-4.67)^ 1.13 (0.96-1.71)
Autonomy 67.1 (10.3) 61.7 (10.8) 60.0 (11.6) 0.59 (0.45-0.76)^ 0.49 (0.37-0.66)^ 0.85 (0.64-1.13)
OR;odds ratio, CI;confidence interval, SD=standard deviation, BAI=beck anxiety inventory, BDI=beck depression inventory, Y BOCS=Yale-Brown obsessi-
Comorbiditiy in OCD
ve-compulsive scale, PI-R: Padua Inventory-Revised

* OR per SD increase using bivariate logistic regression analyses (Mentioned odds ratios and confidence intervals are unadjusted. Significant findings of
adjusted data are mentioned in the article.)
^ p = significant (due to Bonferroni correction significance level was set on 0.002 for clinical factors and on 0.004 for risk indicators)
2.2
Chapter 2.2
Bivariate analyses showed no differences in socio-demographic factors or vulnera-

bility factors. Patients with OCD with current comorbidity had a higher mean score on
the personality trait ‘neuroticism’ and a lower mean score on ‘extraversion’ and ‘auton-
omy’.
After adjusting for all risk indicators that were significant in the bivariate model,
the higher score on neuroticism remained significantly associated with comorbidity
(OR:2.55;95%CI:1.82-3.57, results not shown in Table 2).
We subsequently investigated whether the number of comorbid conditions crit-
ically differentiated pure OCD, OCD with one comorbid disorder and OCD with
multiple comorbid disorders (Table 3). The results of the comparison between pure
OCD (n=172) and OCD with one comorbid disorder (n=119) or with multiple comorbid
disorders (n=91) were similar to the results described above (comparing pure OCD
and OCD with comorbidity (n=210, Table 2) with three exceptions: first, the quality of
life items ‘problems with usual activities’ and ‘pain or discomfort’ were not significant
when comparing pure OCD to OCD with one comorbid disorder. Second, the total
PI-R score was not significant when comparing pure OCD to OCD with one comorbid
disorder. Third, it appeared that OCD with multiple comorbid disorders was also as-
sociated with more severe Y-BOCS symptoms (OR:1.56;CI:1.16-2.11), more chronicity
(OR:2.23;CI:1.26-3.97) and more childhood trauma (OR:3.14;CI:1.53-6.45) compared
to pure OCD. After adjusting for all risk indicators that were significant in the bivariate
model, the higher score on neuroticism remained significantly associated with comor-
bidity among those with one comorbid disorder (OR:2.51;95%CI:1.69-3.72, results not
shown in Table 3), while both neuroticism (OR:2.78;CI:1.78-4.37) and childhood trauma
(OR:4.00;CI:1.77-9.04) remained significantly associated with comorbidity in patients
with multiple comorbidity.
Finally, we compared patients with one comorbid disorder and those with multiple
comorbidity. Multiple comorbidity was associated with higher PI-R scores (OR:1.73;-
CI:1.29-2.30), greater chronicity (OR:2.87;CI:1.56-5.25) and a higher mean BAI score
(OR:2.18;CI:1.60-2.97) and mean BDI score (OR:1.88;CI:1.39-2.55) compared to those
with one comorbid disorder.
Temporal sequencing
To investigate whether OCD precedes or follows comorbidity, three patterns of devel-
opment were investigated in patients with comorbidity. OCD preceded the comorbid
disorder in 45% of the cases, while in 39% of the cases comorbidity preceded the OCD
and in 15% OCD and comorbidity developed simultaneously. Specific differences were
found when comparing OCD patients with one comorbid disorder with OCD patients
with multiple comorbid disorders (Figure 1). Whereas in patients with one comorbid
disorder, OCD preceded the comorbidity in 56% of the cases, in patients with multiple
comorbid disorders comorbidity preceded the OCD in 62%. We found that secondary
OCD was often preceded by another anxiety disorder (66%). Contrasting, in cases of
depressive comorbidity, OCD usually preceded the depression (66%).
When we compared patients with primary OCD to those with primary comorbidity,
no socio-demographic differences were found. The only vulnerability factor that was
significantly more present among patients with primary comorbidity was neuroticism.
52
Comorbiditiy in OCD
Figure 1
Temporal sequencing of OCD and 1 or multiple comorbid disorders (n=168).
2.2
Due to missing data in 42 patients, temporal sequencing was calculated for a subsample of n=168 pa-
tients with comorbidity. An average of 2.6 disorders were present in the ‘multiple comorbidity’ group.
Patients with primary OCD had an earlier age of onset compared to those with second-
ary OCD and more treatment delay. No significant differences were found in terms of
the clinical variables, the severity of the OCD, increased chronicity and more severe
negative consequences on daily life.
DISCUSSION
In this study among OCD patients, comorbidity was at a current rate of 55% and the
lifetime rate was 78%. Comorbid anxiety (lifetime comorbidity 46%) and mood disor-
ders (lifetime comorbidity 64%) were the most frequently found comorbid psychiatric
disorders, in contrast with the low rates of somatoform disorders, substance use disor-
ders, eating disorders and psychotic disorders. Patients with multiple comorbid disor-
ders were more likely to develop comorbidity prior to OCD, while patients with one co-
morbid disorder were more likely to develop comorbidity after the development of OCD.
The comorbidity rates found in this study confirm the high rates of lifetime and cur-
rent comorbidity in OCD that were previously reported.4,5 This does not hold for the
lower comorbidity rates of anxiety disorders in one of these studies.5 The authors in that
study account for these rather low comorbidity rates by stating that their sample was
drawn from an academic department that specializes in anxiety disorders, which may
have caused an underestimation of comorbidity.5 By contrast, in the current study and in
Pinto’s study4 both inpatients and outpatients were recruited from various settings such
as a private psychiatric hospital, a specialist OCD clinic and general psychiatric clinics.
53
Chapter 2.2
Clinical characteristics of comorbidity

We found that the severity of OCD did not differ substantially between ‘pure’ OCD and
OCD with comorbid disorders. However, when breaking down the group of comorbid
disorders in OCD with one comorbid disorder and OCD with two or more comorbid
disorders, it appeared that the latter group of patients suffered from significantly more
severe OCD and more chronic OCD. The same pattern was found for an association
between comorbidity and OCD chronicity; when differentiating between OCD and one
or multiple comorbid disorders, it appeared that multiple comorbidity is associated
with more chronicity. These results may explain the inconsistencies in the literature
hitherto; whereas some studies did not find any association between comorbidity and
OCD severity5 or chronicity,8 other studies did find an association between comorbid-
ity and OCD severity11 or chronicity.6 A possible explanation could be the differenti-
ation between one and multiple comorbid disorders that we made, as did those two
studies.6,11
Comorbidity was associated with more anxiety and depressive symptoms, more
negative consequences in daily life, but not with an earlier age of onset or treatment
delay. These findings confirm earlier findings.5,11 In our sample, comorbidity was relat-
ed to the specific symptom dimensions of checking (according to the Y BOCS), and
precision, rumination and impulses (Pi-R). The meaning of these findings has yet to
be determined while the use of symptom dimensions is not unambiguous and earlier
research is inconsistent: while Hasler and colleagues35 found an association between
depressive comorbidity and aggressive and sexual obsessions, Quarantini and col-
leagues36 could not replicate this.
To sum up, patients with OCD and a comorbid disorder suffer from greater chro-
nicity, more severe OCD and more negative consequences on their daily life, but they
are not treated differently from patients with pure OCD. What does this imply for clinical
practice? Frequently, pure OCD and OCD with comorbidity are being treated in the
same way. We believe our findings recommend taking into account comorbidity when
planning treatment; a comorbid severe depression (even secondary to OCD) may be
an indication for antidepressants, while the treatment of a primary anxiety disorder
with a secondary OCD should not focus only on this primary anxiety disorder, but on
OCD as well. This contrasts with the current opinion about comorbidity in other anxiety
disorders (e.g. panic disorder and generalized anxiety disorder) in which the severity of
comorbidity usually declines when the primary diagnosis is treated successfully37,38 and
caution is advised in relation to the transdiagnostic treatment of comorbid disorders.39
Finally, we found a different pattern of comorbidity when we compared primary and
secondary OCD; in the case of depressive comorbidity, OCD is often developed prior
to the mood disorder, while secondary OCD is usually preceded by another anxiety dis-
order. However, the temporal sequencing results may be a coincidental finding. If rep-
licated, their clinical implications have yet to be determined. We hypothesize two types
of OCD; a more endogenous form (primary OCD) which is a risk factor for developing
other types of psychiatric disorders, and a more secondary form of OCD that develops
as a coping mechanism in relation to another anxiety disorder.
54
Comorbiditiy in OCD
Risk indicators for comorbidity

Childhood trauma and neuroticism are vulnerability factors associated with comor-
bidity in OCD, especially with multiple comorbidity. Although instances of traumatic
experience occurred in higher proportions in OCD than in the normal population, these
findings are not specific to OCD, since both high rates of neuroticism40-42 and childhood
trauma43-45 have been associated with comorbidity in depressive disorder40,44, schizo-
phrenia,41,45 and post-traumatic stress disorder.42 Neuroticism seems to reflect general
vulnerability, and childhood trauma seems to lead to a general vulnerability and neither
is therefore useful as specific predictors of comorbidity in OCD. However, these gener-
al predictors, combined with the fact that patients with multiple disorders often develop
OCD secondary to the comorbidity, imply that clinicians should be aware of OCD in
complex patients with childhood trauma and multiple disorders.
2.2
Strengths and limitations
The current study adds to current knowledge since it uses a large, well-defined clinical
sample of 382 current OCD patients in settings that varied between general psychi-
atric outpatient clinics to specialist OCD inpatient clinics. The patients included can
therefore be generalized to clinical practice. The use of the SCID-I/P as a diagnostic
instrument and of validated instruments to study clinical factors, consequences, vulner-
ability factors and personality structure guarantee the statistical power, reliability and
validity of the results. However, it should be noted that causal inferences about asso-
ciation between risk factors and OCD are precluded since our results were based on
cross-sectional data. Prospective data are needed to clarify causality in the pathways
of comorbidity. Furthermore, the temporal sequencing and childhood trauma data were
collected retrospectively, which may have introduced some recall bias.
This study did not investigate the possible differential impact of types of comorbid
disorders on OCD, but instead focused only on the number of comorbid disorders due
to a lack of power. However, this may be an important variable for future studies.
Finally, we chose to break down the group of patients with OCD and comorbidity
into ‘one comorbid disorder’ and ‘multiple comorbid disorders’ to achieve sufficient
statistical power. This decision was rather arbitrary since it was not based on earlier
findings in OCD. However, there is evidence in other anxiety disorders that multiple co-
morbidity is clinically relevant.46
Conclusion
The results presented here show that comorbidity rates in OCD are high. Secondary
OCD is often preceded by an anxiety disorder while primary OCD often precedes a
mood disorder. Comorbidity, especially of multiple comorbid disorders, is clinically rel-
evant since it is associated with a specific vulnerability pattern, with greater chronicity,
more severe OCD and more negative consequences on daily life. This indicates that
the diagnosis and treatment of all comorbid disorders is clinically relevant, and that
clinicians should be especially aware of multiple disorders in patients with childhood
trauma and high neuroticism scores. Further research should focus on the long-term
consequences of comorbidity.
55
Chapter 2.2
Acknowledgements
The research infrastructure needed to complete the baseline measurements (including
personnel and materials) is financed almost exclusively by the participating organisa-
tions: Academic department VU Medical centre/GGZinGeest Amsterdam, The Nether-
lands; Marina de Wolf anxiety research centre, GGZ Centraal, Ermelo; Center for Anx-
iety Disorders “Overwaal”, Lent; Dimence, GGZ Overijssel; Department of Psychiatry,
Leiden University Medical Centre Leiden; ‘Vincent van Gogh institute‘ mental health
care centre Noord- en Midden-Limburg, Venray; Academic Anxiety Center, PsyQ
Maastricht/Maastricht University, Division Mental Health and Neuroscience), except for
the fieldwork coordinator, who was financed for the duration of one year by a research
grant from the Stichting Steun.
56
Comorbiditiy in OCD
Reference list
1. Steketee G, Eisen J, Dyck I et al. Predictors of course in Obsessive Compulsive
disorder. Psychiatry Res 1999;89:229-238.
2. Weissman MM, Bland RC, Canino GJ et al. The cross national epidemiology of ob-
sessive compulsive disorder. J Clin Psychiatry 1994;55:5-10.
2001;110:585-599.
try 2006;67:703-711.
ple of patients with Obsessive compulsive disorder. J Affect Disord 2004;80:155-
162.
2.2
6. Jakubovski E, Diniz JB, Valerio C et al. Clinical predictors of long-term outcome in
obsessive-compulsive disorder. Depress Anxiety 2012;00:1-10.
7. Frost RO, Steketee G, Williams LF et al. Mood, personality disorder symptoms
and disability in obsessive compulsive hoarders: a comparison with clinical and
non-clinical controls. Behav Res Ther 2000;38:1071-1081.
8. Eisen JL, Goodman WK, Keller MB et al. Patterns of remission and relapse in
obsessive-compulsive disorder: a 2-year prospective study. J Clin Psychiatry
1999;60:346-351.9.
9. Zitterl W, Demal U, Aigner M et al. Naturalistic Course of Obsessive Compulsive
Disorder and Comorbid Depression. Longitudinal Results of a Prospective Fol-
low-up Study of 74 Actively Treated Patients. Psychopathology 2000; 33:75-80.
10. Rasmussen SA, Eisen JL. Clinical and epidemiologic findings of significance to
neuropharmacologic trials in OCD. Psychopharmacol Bull 1988;24:466-470.
11. Tükel R, Polat A, Özdemir Ö et al. Comorbid Conditions in Obsessive-Compulsive
Disorder. Compr Psychiatry 2002;43:204-209.
12. Başoğlu M, Lax T, Kasvikis Y et al. Predictors of Improvement in Obsessive-Com-
pulsive Disorder. J Anxiety Disord 1988;2:299-317.
13. Lensi P, Cassano GB, Correddu G et al. Obsessive-Compulsive Disorder: Famil-
ial-Developmental History, Symptomatology, Comorbidity and Course with Special
Reference to Gender-Related Differences. Br J Psychiatry 1996;169:101-107.
14. Ruscio AM, Stein DJ, Chiu WT et al. The epidemiology of obsessive-compul-
sive disorder in the National Comorbidity Survey Replication. Mol Psychiatry
2010;15:53-63.
15. Schuurmans J, van Balkom AJ, van Megen, HJ et al. The Netherlands OCD Asso-
chiatr Res 2012;21:273-285.
16. First MB, Spitzer RL, Gibbon M et al. Structured clinical interview for DSM-IV-TR
axis I disorders, patient Edition (SCID-I/P, version 2.0). New York: Biometrics Re-
search, New York States psychiatric institute;1996.
17. Lobbestael J, Leurgans M, Arntz A. Inter-Rater Reliability of the Structured Clinical
57
Chapter 2.2
Interview for DSM-IV Axis I Disorders (SCID-I) and Axis II Disorders (SCID-II).
Clinical Psychol Psychother 2011;18:75-79.
18. van Oppen P, Hoekstra RJ, Emmelkamp PM. The structure of obsessive-compul-
sive symptoms. Behav Res Ther 1995;33:15-23.
19. Goodman WK, Price LH, Rasmussen SA et al. The Yale-Brown Obsessive
Compulsive Scale: I. Development, use, and reliability. Arch Gen Psychiatry
1989;46:1006–1011.
20. Sanavio E. Obsessions and Compulsions; the Padua Inventory. Behav Res Ther
1988;26:167-177.
21. Goodman WK, Price LH, Rasmussen SA et al. The Yale–Brown obsessive com-
pulsive scale. II. Validity. Arch Gen Psychiatry 1989;46:1012–1016.
22. Leckman JF, Grice DE, Boardman J et al. Symptoms of Obsessive-Compulsive
Disorder. Am J Psychiatry 1997;154:911-917.
23. Bear L. Factor analysis of symptom subtypes of obsessive compulsive disorder
and their relation to personality and tic disorders. J Clin Psychiatry 1994;55:18-23.
24. Katerberg H, Delucchi KL, Stewart SE et al. Symptom dimensions in OCD:
item-level factor analysis and heritability estimates. Behav Genet 2010;40:505-517.
25. Anholt GE, Aderka IM, van Balkom AJ et al. Age of onset in OCD: Admixture anal-
ysis with a large sample. Psychological medicine 2014;44:185-194.
26. Beck AT, Ward CH, Mendelson M et al. An inventory for measuring depression.
Arch Gen Psychiatry 1961;4:561-571.
27. Beck AT, Epstein N, Brown G et al. An inventory for measuring clinical anxiety:
psychometric properties. J Consult Clin Psychol 1988;56:893-897.
29. Richter P, Werner J, Heerlein A et al. On the Validity of the Beck Depression Inven-
tory. A review. Psychopathology 1998;31:160-168.
30. EuroQol Group. EuroQol- a new facility for the measurement of health-related
quality of life. Health Policy 1990;16:199-208.
31. Kind P, Dolan P, Gudex C et al. Variations in population health status: Results form
a United Kingdom national questionnaire survey. BMJ 1998;316:736-741.
32. Draijer N. Structured Trauma Interview. Amsterdam, Vrije Universiteit, Department
of Psychiatry:1989.
33. Hendriks AA, Hofstee WK, De Raad B. The Five-Factor Personality Inventory
(FFPI). Pers Individ Dif 1999;27:307-325.
34. Hendriks AA, Perugini M, Angleitner A et al. The Five-Factor Personality Inventory:
Cross-cultural generalizability across 13 countries. Eur J Pers 2003;17:347-373.
35. Hasler G, Holland LaSalle-Ricci V. Ronquillo JG et al. Obsessive-compulsive dis-
order symptom dimensions show specific relationships to psychiatric comorbidity.
Psychiatry Research 2005;135:121-132.
36. Quarantini LC, Torres AR, Sampaio AS et al. Comorbid major depression in obses-
sive-compulsive disorder patients. Comprehensive Psychiatry 2011;52:386-393.
37. Tsao JC, Mystkowski JL, Zucker BG et al. Effects of cognitive-behavioral therapy
for panic disorder on comorbid conditions: Replication and extension. Behav Res
Ther 2002;33:493–509.
58
Comorbiditiy in OCD
38. Borkovec TD, Abel JL, Newman H. Effects of psychotherapy on comorbid condi-
tions in generalized anxiety disorder. J Consult Clin Psychol 1995;63:479-483.
sion. Depress Anxiety 2012;29:749-753.
40. Enns MW, Cox BJ. Personality dimensions and depression: review and commen-
tary. Can J Psychiatry 1997;42:274-84.
41. Van Os J, Jones PB. Neuroticism as a risk factor for schizophrenia. Psychol Med
2001;31:1129–1134.
42. Cox BJ, MacPherson PS, Enns MW et al. Neuroticism and self-criticism associat-
ed with posttraumatic stress disorder in a nationally representative sample. Behav
Res Ther 2004;42:105–114.
43. Kessler RC, Davis CG, Kendler KS. Childhood adversity and adult psychiatric dis-
order in the US National Comorbidity Survey. Psychol Med 1997;27:1101-1119.
44. Hovens JG, Wiersma JE, Giltay EJ et al. Childhood life events and childhood trau-
2.2
ma in adult patients with depressive, anxiety and comorbid disorders vs. controls.
Acta Psychiatric Scand 2010;122:66-74.
45. Read J, van Os J, Morrison AP et al. Childhood trauma, psychosis and schizophre-
nia: a literature review with theoretical and clinical implications. Acta Psychiatric
Scand 2005;112:330-350.
46. Hofmeijer-Sevink MK, Batelaan NM, van Megen HJ et al. Clinical relevance of co-
morbidity in anxiety disorders: A report from the Netherlands Study of Depression
and Anxiety (NESDA). J Affect Disorder 2012;137:106-112.
59
60
Chapter 2.3
Occurrence and predictive value of clinically relevant

obsessive-compulsive symptoms in anxiety and depressive
disorders
Mieke Klein Hofmeijer-Sevink, Neeltje M. Batelaan, Harold J.G.M. van Megen,

Marcel A. van den Hout, Brenda W. Penninx, Anton J.L.M. van Balkom, Danielle
C. Cath.
Submitted for publication
Author Contributions: All of the authors were involved in the conception and design of this study and/or in
the analysis and interpretation of data for this manuscript. All of the authors drafted the manuscript and/or con-
tributed important intellectual content to its revision.
61
Chapter 2.3
Abstract
Background: Comorbidity between anxiety and depressive disorders is common, and
associated with higher illness severity and poorer treatment outcome. However, Obses-
sive Compulsive Disorder (OCD) is generally not assessed in epidemiological studies
and consequently, whether comorbid obsessive compulsive symptoms impact on clini-
cal outcomes in anxiety and depressive disorders is unknown.
Methods: Two year follow up data were obtained from the Netherlands Study of Anx-
iety and Depression (NESDA), including a total sample of 2125 healthy controls and
participants with a remitted anxiety and/or depressive disorder, a depressive disorder,
an anxiety disorder, and with a combined anxiety and depressive disorder. The Young
Adult Self-Report Scale for obsessive compulsive symptoms (YASR-OCS) was used
to assess clinically relevant OC-symptoms. We examined the occurrence and severity
of OC-symptoms, the association with anxiety and/or depressive disorders; and the
predictive value of OC-symptoms on the two year course trajectory of anxiety and/or
depressive disorders.
Results: Clinically relevant OC-symptoms occurred in 35.9-66.5% of the participants
with an anxiety and/or depressive disorder and were associated with severity. Pres-
ence of clinically relevant OC-symptoms predicted first onset (OR 5.79, 95% CI 1.15-
29.14), relapse (OR 2.31, 95% CI 1.55-3.46), and persistence in those with comorbid
anxiety and depressive disorders (OR 4.42, 95%CI 2.54-7.70).
Conclusions: Clinically relevant OC-symptoms (and probably OCD) are present in
a substantial proportion of patients with anxiety and/or depressive disorders and are
associated with higher illness severity and unfavourable course trajectories. These
results indicate that clinicians should be aware of significant OC-symptoms in anxiety
and depressive disorder, and moreover, that these OC symptoms warrant treatment, to
counter-act these unfavourable outcomes.
62
OC symptoms in anxiety and depression
INTRODUCTION
Both anxiety and depressive disorders are clinically heterogeneous and individual pa-
tients differ with respect to severity, prognosis and treatment response. Comorbidity
between anxiety and depressive disorders is common, with rates estimated between
40-80%1 and it is often associated with higher illness severity, impaired functioning and
poorer treatment outcome.2-6 However, Obsessive Compulsive Disorder (OCD) is gen-
erally not assessed in epidemiological studies. Consequently, it is currently unknown
whether comparable unfavourable outcomes apply to the comorbidity between OCD
and anxiety and depressive disorders.
The relatively sparse research that assesses obsessive compulsive symptoms
(OC-symptoms) in anxiety and depressive disorders shows that among patients with
anxiety or depressive disorders, OC-symptoms occurs in 30-40% of patients.4,7 Howev-
er, a systematic and direct examination of the impact of OC-symptoms on clinical out-
comes in anxiety and depressive disorders is currently lacking. Therefore it is unknown
whether comorbidity with OC-symptoms would lead to increased severity of anxiety
and depression symptoms and to chronicity of the illnesses.
Using longitudinal data from the Netherlands Study of Depression and Anxiety 2.3
(NESDA), the present study was undertaken to examine the clinical relevance of co-
morbid OC-symptoms in the context of anxiety and depressive disorders. This study
aims to i) investigate the occurrence and severity of OC-symptoms in a large sample
of healthy controls and participants with anxiety or depressive disorders and; ii) in-
vestigate the association between clinically relevant OC-symptoms and the presence
and severity of anxiety and/or depressive disorders, iii) examine the predictive value of
clinically relevant OC-symptoms with respect to first onset, relapse and persistence of
anxiety and/or depressive disorders, and to examine whether this association is inde-
pendent of severity of anxiety and/or depressive symptoms.
MATERIALS AND METHODS

Procedure
Data were obtained from the Netherlands Study of Depression and Anxiety (NESDA).
NESDA encompasses a longitudinal cohort study that has included 2981 adults (18-65
years) at baseline with anxiety disorders (panic disorder, social phobia, generalized
anxiety disorder (GAD) and agoraphobia), with depressive disorders (major depres-
sive disorder, dysthymic disorder), and healthy controls. A detailed description of the
NESDA study design can be found elsewhere.8 In short: the NESDA study aims to
describe the long-term course and consequences of depressive and anxiety disor-
ders in a sample recruited from the community, primary care settings and specialized
mental health care facilities. Persons with a clinically overt primary diagnosis of other
psychiatric conditions, such as psychosis, post-traumatic stress disorder and obses-
sive-compulsive disorder were excluded from the study, as were persons not fluent in
Dutch language. The Young Adult Self-Report Scale for obsessive compulsive symp-
toms9 (YASR-OCS) was administered at two year follow up wave, allowing to study
OC-symptoms.
Approval of the study was granted by the Ethical Review Boards of all participating
centers and written informed consent was obtained from all participants.
63
Chapter 2.3
Approval of the study was granted by the Ethical Review Boards of all the partici-
pating centers and written informed consent was obtained from all participants.
Sample
OC-symptoms were first assessed at the two year follow up. We therefore consider this
wave as baseline for the present study, and the four year wave of NESDA as our two
year follow up. For the present study all participants who enrolled in both the two year
and the four year follow up (n=2402) were included. Attrition between baseline and the
four year follow up (n=579) was not related to gender and age, but related to education
and the presence of an anxiety and/or depressive disorder at baseline.
Participants who had not returned the YASR-OCS questionnaire (n=270) and those
with too many missing data (n=7) were excluded. Thus, the final sample consisted of
2125 participants. Those who were included (n=2125) were older, more frequently fe-
male, less educated and had more anxiety and/or depressive disorders at baseline as
compared to those who were excluded (n=277).
Figure 1
Figure
Items 1. Items
of the ofAdult
Young the Self-Report
Young AdultScale, Self-Report Scale,symptoms
obsessive-compulsive obsessive-compulsive
(YASR-OCS,
Achenbach, 1997).
symptoms (YASR-OCS, Achenbach, 1997).
Item 1 ‘I Cannot get my mind off certain thought’
Item 2 ‘I repeat certain acts over and over’
Item 3 ‘I am afraid I might think or do something bad’
Item 4 ‘I have thoughts that other people think are strange’
Item 5 ‘I do things other people think are strange’
Item 6 ‘I feel I have to be perfect’
Item 7 ‘I worry a lot’
Item 8 ‘I feel too guilty’
Measures
OC-symptoms
To assess presence and severity of OC-symptoms, the 8-item Young Adult Self-Report
Scale for obsessive compulsive symptoms (YASR-OCS) was used, which is a sub-
scale of the YASR9 (see Figure 1 for items). Participants respond on a 3-point scale
(0, not true, 1, somewhat true or sometimes true, 2, very true or often true), yielding
score ranges between 0-16 on the YASR-OCS. The internal reliability of the YASR-
OCS in our sample was good (Cronbach’s alpha 0.83). Scoring ≤7 versus >7 has been
64
previously identified as the best cut-off to predict OCD, with a sensitivity of 82.4% and
specificity of 69.7% when compared to clinical controls.10 Since the YASR-OCS is a
screening instrument and not a DSM-IV diagnostic tool, participants with scores above
the cut-off point of 7 were regarded as suffering from clinically relevant OC-symptoms
(and not definite OCD).
Diagnostic status. To diagnose anxiety and depressive disorders, the CIDI version
2.1 was used, using DSM-IV criteria. The CIDI is a structured interview with acceptable
reliability and validity11,12 and was administered by trained research staff.
Healthy controls were defined as participants without a lifetime history of a CIDI
diagnosis of an anxiety or depressive disorder. The group defined as “remitted” con-
tained participants with a lifetime history of an anxiety and/or depressive disorder but
no anxiety or depressive disorder in the past six months. The group defined as “de-
pressive disorder” and as “anxiety disorder” contained participants with a CIDI diag-
nosis of a major depressive disorder and/or dysthymia, and of a panic disorder, social
phobia, generalized anxiety disorder, and/or agoraphobia respectively in the past six
months. The group defined as “combined anxiety and depressive disorder” contained
participants with both an anxiety and a depressive disorder in the past six months 2.3
Covariates. All covariates were assessed at the ‘baseline’ of the present study (i.e.
two year follow up of the NESDA study).
Socio-demographic factors included gender (male/female), age and years of educa-
tion.
Symptom severity of anxiety and depression were assessed to examine whether
high YASR-OCS scores 1) were associated with severity of anxiety and/or depressive
disorders and 2) predicted course trajectory independently of anxiety and depression
severity. Severity of anxiety symptoms was assessed with the Beck Anxiety Inventory
(BAI).13 The BAI is a widely used 21-item self-report questionnaire which has proven to
be valid and reliable.14 To determine the severity of depressive symptoms the Inventory
of Depressive Symptomatology (IDS) was used.15 This scale is a 30-item self-report
questionnaire measuring depressive symptoms which has proven to be valid and reli-
able.16
Course trajectory were measured in three different ways for specific subsamples; i)
In healthy controls, first onset was defined as six-months presence of an anxiety and/or
depressive disorder at follow up. ii): In participants with a remitted disorder at baseline,
relapse was defined as six-months presence of an anxiety and/or depressive disorder
at follow up. iii) In participants with an anxiety and/or depressive disorder at baseline,
persistence was defined as six-months presence of the same diagnosis at follow up.
Statistical analyses
To compare participants with high versus low YASR-OCS scores on socio-demograph-
ic factors, symptom severity and course trajectory 2-tailed Chi-square statistics were
used for categorical variables and 1-way analysis of variance statistics (ANOVA) for
continuous variables. Bivariate and multivariate logistic regression analyses were used
to evaluate the influence of YASR-OCS scores on the onset and course of anxiety and/
or depressive disorders. First, bivariate logistic regression analyses were conducted
and second, multivariate logistic regressions analyses were conducted, adjusting for all
65
Chapter 2.3
covariates that were significant in the bivariate analyses. All statistical analyses were
conducted with the SPSS statistical package version 20.0 (SPSS Inc., Chicago, Illi-
nois). P-values <0.05 were considered significant.

Table 1
Baseline characteristics of participants with low- (n=1624) versus high (n=501) YASR-OCS scores
total sample low YASR-OCS high YASR-OCS χ2/ F p

(n=2125) (n=1624, 76.4%) (n=501, 23.6%) (df)
Socio-demographic
factors
Age (mean, SD) 44.6 (13.2) 44.9 (13.2) 43.4 (12.9) 5.4 (1) 0.02
Gender (% female, n) 66.7 (1418) 67.9 (1102) 63.1 (216) 3.9 (1) 0.05
Education (mean 12.7 (3.3) 12.7 (3.3) 12.4 (3.3) 4.5 (1) 0.04
number of years, SD)
Severity
IDS (mean, SD) 15.3 (11.8) 11.7 (9.3) 26.6 (11.8) 862.7 <0.0001
BAI (mean, SD) 8.3 (8.4) 6.2 (6.7) 15.2 (9.6) 556.5 <0.0001
YASR-OCS: Young Adult Self-Report Scale, obsessive-compulsive symptoms (Achenbach, 1997).

Low YASR-OCS:≤ 7; High YASR-OCS:> 7.
IDS: Inventory of Depressive Symptomatology (Rush et al., 1996). BAI: Beck Anxiety Inventory
(Beck et al., 1988). SD: standard Deviation
RESULTS
Sample description
The sample consisted of 2125 participants with an average age of 44.6 years (SD 13.2),
who were predominantly female (66.7%) and had an average of 12.7 years of education
(SD 3.3). The mean IDS score was 15.3 (SD 11.8) and the mean BAI score was 8.3 (SD
8.4) (see Table 1). Of the total sample, 22.1% (n=469) were healthy controls, 42.8%
(n=909) had a remitted anxiety and/or depressive disorder, 9.6% (n=205) had a current
depressive disorder, 12.7% (n=270) had a current anxiety disorder and 12.8% (n=272)
had a current combined anxiety and depressive disorder (not shown in Table 1).
Occurrence and severity of obsessive compulsive symptoms

Using the predefined threshold on the YASR-OCS, 76.4% of the sample (n=1624) had
few (clinically not relevant) OC-symptoms (≤ 7), whereas the remaining 23.6% (n=501)
had high (clinically relevant) OC-symptoms (>7). Participants with high YASR-OCS
66
scores were more likely to be male, younger and reported less years of education. In
addition, high YASR-OCS scores were associated with higher anxiety and depressive
symptoms (See Table 1).

Figure 2. Distribution of low versus high YASR-OCS scores across healthy
Figure 2
controls, participants
Distribution with
of low versus high a remitted
YASR-OCS anxiety
scores across and/orparticipants
healthy controls, depressive disorder and
with a re-
mitted anxiety and/or depressive disorder and participants with a current anxiety and/or depressive
participants with a current anxiety and/or depressive disorder (n=2125).
disorder (n=2125).
120
97,9
100
84,7
80 66,5
63,9 64,1
60
36,1 35,9
2.3
40 33,5
low YASR-OCS
15,3
20
2,1 high YASR-OCS
0
healthy remitted current current current
controls anxiety depressive anxiety anxiety
(n=469) and/or disorder disorder and
depressive (n=205) (n=270) depressive
disorder disorder
(n=909) (n=272)

YASR-OCS: Young Adult
Low YASR-OCS:≤ Self-Report
7; High YASR-OCS:> Scale,
7. obsessive-compulsive symptoms (Achenbach, 1997).
Overall statistics: χ2:473.74(4),p<0.001.
Overall statistics: χ2:473.74(4),p<0.001.
Association between clinically relevant OC-symptoms and presence and severi-

ty of anxiety and/or depressive disorders
Low YASR-OCS scores were predominantly found in healthy controls and in those
with remitted anxiety and/or depressive disorders, whereas high YASR-OCS scores
were predominantly found in those with an anxiety and/or depressive disorder. Thus,
high YASR-OCS scores were associated with more anxiety and depression (χ2: 473.74
(4), p<0.001). Results are presented in Figure 2. More specifically, among participants
with a depressive disorder, high YASR-OCS scores were associated with more severe
depressive symptoms: participants with high YASR-OCS had mean IDS scores of 26.4
(SD 11.0), versus mean IDS scores of 20.3 (SD 10.8; F:15.07 (1,203), p<0.001) in those
with low YASR-OCS. Likewise, in those with an anxiety disorder, high YASR-OCS
67
Chapter 2.3
scores were associated with more severe anxiety symptoms: participants with high
YASR-OCS had mean BAI scores of 14.5 (SD:8.5) versus mean BAI scores of 11.8
(SD 8.7; F:5.95 (1,268), p:0.02) in those with low YASR-OCS. Finally, in patients with a
combined anxiety and depressive disorder, high YASR-OCS scores were associated
with more severe depressive and anxiety symptoms; participants with high YASR-OCS
had mean IDS scores of 33.7 (SD:11.5) and mean BAI scores of 20.3 (SD:10.0), versus
mean IDS scores of 24.7 (SD 10.7;F:39.86 (1,270), p<0.001) and mean BAI scores of
13.4 (SD:8.4; F:31.22 (1,270), p<0.001) in patients with low YASR-OCS scores.

Table 2
Impact of high YASR-OCS scores on the two year course trajectory of anxiety and depressive disor-
ders (n=2125)
Course trajectory low YASR-OCS High YASR-OCS χ2 (df) P

(%, n) (%, n)
First onset in healthy controls (n=469)

First onset (n=21) 90.5 (19) 9.5 (2) 5.76 (1) 0.02
No first onset (n=448) 98.2 (440) 1.8 (8)
Relapse in participants with a remitted

anxiety and/or depressive disorder
(n=909)
Relapse (n=177) 74.6 (132) 25.4 (45) 17.42 (1) <0.0001
No relapse (n=732) 87.2 (683) 12.8 (94)
Persistence in participants with a

depressive disorder (n=205)
Persistence (n=84) 59.5 (50) 40.5 (34) 1.18 (1) 0.28
No persistence (n=121) 66.9 (81) 33.1 (40)
Persistence in participants with an

anxiety disorder (n=270)
Persistence (n=138) 63.8 (88) 36.2 (50) 0.01 (1) 0.92
No persistence (n=132) 64.4 (85) 35.6 (47)
Persistence in participants with an

anxiety and depressive disorder (n=272)
Persistence (n=192) 23.4 (45) 76.6 (147) 15.06 (1) <0.0001
No persistence (n=80) 57.5 (46) 42.5 (34)

Predictive value of clinically relevant OC-symptoms on first onset and two year
course trajectories of anxiety and depressive disorders
First onset of anxiety and depressive disorders at two year follow up was present in
4.5% (n=21) of the healthy controls (see Table 2).
68
Table 3
Predicted value of clinically relevant-OC symptoms for first onset, relapse and persistence of anxiety and/or depressive disorders (n=2125)
First onset in Relapse in remitted anxiety Persistence Persistence Persistence in combined anxiety
healthy controls and/or depressive disorders in depressive in anxiety and depressive disorder
disorder disorder
(n=469) (n=909) (n=205) (n=270) (n=272)
69
Model 1 Model 1 Model 2 Model 1 Model 1 Model 1 Model 2
OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
OC-symptoms
High YASR-OCS 5.79 (1.15-29.14) 2.31 (1.55-3.46) 1.37 (0.88-2.14) 1.38 (0.77-2.45) 1.03 (0.63-1.69) 4.42 (2.54-7.70) 3.36 (1.86-6.07)
Symptom severity
IDS* 2.22 (1.77-2.28) 2.01 (1.46-2.77) 1.98 (1.43-2.74) 1.33 (0.99-1.89) 1.73 (1.30-2.29) 1.10 (0.74-1.66)
BAI* 1.75 (1.42-2.16) 1.05 (0.78-1.41) 1.51 (1.23-2.04) 1.54 (1.19-2.00) 1.71 (1.32-2.22) 1.38 (0.96-1.98)
Socio-demographic s
Age* 1.07 (0.91-1.26) 1.05 (0.87-1.45) 0.97 (0.75-1.23) 1.09 (0.82-1.43)
Female gender 1.21 (0.85-1.73) 0.89 (0.48-1.64) 1.06 (0.63-1.81) 0.74 (0.42-1.30)
Education* 0.95 (0.80-1.12) 0.95 (0.72-1.29) 0.93 (0.73-1.18) 0.84 (0.85-1.08)
YASR-OCS: Young Adult Self-Report Scale, obsessive-compulsive symptoms (Achenbach, 1997). Low YASR-OCS:≤ 7; High YASR-OCS:> 7.
IDS: Inventory of Depressive Symptomatology (Rush et al., 1996). BAI: Beck Anxiety Inventory (Beck et al., 1988).
OR: odds ratio, CI: confidence interval. * OR presented per SD increase.
Model 1: bivariate logistic regression analyses, model 2: multivariate logistic regressions analyses, adjusted for all covariates significant in model 1.
2.3
Chapter 2.3
Relapse within two years occurred in 19.5% (n= 177) of the participants with a re-
mitted disorder at baseline and this was significantly associated with high YASR-OCS
scores (see Table 2).
Persistence of the baseline disorder was common and occurred in 41.0% (n=84) of
participants with a depressive disorder, in 51.1% (n= 138) of those with anxiety disor-
ders and in 70.6 % (n=192) of the participants with a combined anxiety and depressive
disorder. Persistence was significantly associated with high YASR-OCS scores in
participants with a combined anxiety and depressive disorder, but not in those with an
anxiety disorder or depressive disorder.
Table 3 presents the association between high YASR-OCS scores, socio-demo-
graphic factors, symptom severity and the presence of an anxiety and/or depressive
disorder at two year follow up. In healthy controls with a first onset of an anxiety and/
or depressive disorder (n=21), only two reported clinically relevant OC-symptoms (OR
5.79, 95% CI 1.15-29.14, see Table 3). These low numbers impeded further multivariate
analyses. Post hoc analyses showed that these 21 participants had significant higher
mean YASR-OCS scores at baseline compared to the healthy controls who did not
develop an anxiety and/or depressive disorder (mean YASR-OCS score 3.1 versus 1.4;
F:18.02(1,467), p<0.0001), suggesting that OC-symptoms in healthy controls are asso-
ciated with first onset of anxiety or depressive disorder at two year follow up.
In participants with a remitted disorder, high YASR-OCS scores were significantly
associated with relapse (OR 3.10, 95% CI 2.13-4.50), but the association lost signifi-
cance after adjusting for severity of anxiety and depressive symptoms.
High YASR-OCS scores were not significantly associated with persistence in de-
pressive disorders (OR 1.38, 95% CI 0.77-2.45) or anxiety disorders (OR 1.03, 95% CI
0.63-1.69), but high YASR-OCS scores were significantly associated with persistence
in combined anxiety and depressive disorders (OR 4.42, 95% CI 2.54-7.70).This finding
remained significant when adjusting for severity of anxiety and depressive symptoms
(OR 3.36, 95% CI 1.86-6.07).

DISCUSSION
In this exploratory study on the relationship between OC-symptoms and anxiety and
depressive disorders, we found that clinically relevant OC-symptoms were present in
a substantial proportion of participants, mostly in those with anxiety disorders (35.9%),
depressive disorders (36.1%) and the combination of anxiety and depressive disorders
(66.5%). Further, in participants with an anxiety and/or depressive disorder, the pres-
ence of clinically relevant OC-symptoms was associated with higher symptom severity
of the disorders.
Additionally, clinically relevant OC-symptoms predicted first onset of anxiety and/or
depressive disorder, relapse of anxiety and/or depressive disorders, and persistence in
those with a combined anxiety and depressive disorder, but not persistence in partici-
pants with an anxiety or depressive disorder. The predictive value was not independent
of severity of anxiety and depression, except in participants with a combined anxiety
and depressive disorder.
70
Clinically relevant OC-symptoms were present in a substantial proportion of patients

with anxiety and/or depressive disorders. Moreover, the cut-off score of the YASR-OCS
used in the present study, provides a sensitivity of 82.4% to predict presence of obses-
sive-compulsive disorder.10 Results might thus indicate that OCD is quite common. On
the other hand, the YASR-OCS has a relatively high change of ‘false positive’ findings
(given the specifity of 69.7%).10
Community-based studies in the USA have estimated one year prevalence of OC
symptoms (without meeting full DSM-IV criteria) in the general population at 20-25%,7
and in the Netherlands, one year prevalence of OCD in the general population has
been estimated at 0.5%.17 As compared to these studies, our clinical sample might con-
tain higher levels of OC symptoms, as a result of recruitment in primary and secondary
health care institutions. In sum, although the prevalence rates found in this clinical
study seem somewhat inflated, and difficult to compare to the community-based stud-
ies, clinically relevant OC-symptoms frequently seem to co-occur in populations with
anxiety and depressive disorders and might be easily missed since patients do not
report these symptoms by themselves.18 Clinicians should proactively ask for the pres-
ence of these symptoms, as these may warrant treatment, and when treated, the rela- 2.3
tively unfavourable course of patients with OC symptom comorbidity might be bended
into a more favourable outcome.
No prior research has systematically examined the impact of comorbid OC-symp-
toms on clinical outcomes in anxiety and depressive disorders. However, previous
research from a somewhat different angle investigating i) the impact of anxiety or de-
pressive comorbidity on severity and course of OCD, and ii) the impact of comorbidity
(other than OCD) in anxiety and depressive disorders reported that comorbidity was
associated with severity and chronicity.3-5,19,20 This is consistent without our findings,
showing that the presence of clinically relevant OC-symptoms is associated with sever-
ity of anxiety and/or depressive disorders, and with unfavourable course trajectories.
We found that relapse occurred in one fifth of our sample, and that clinically relevant
OC-symptoms predicted relapse. Given the high rate of relapse in both anxiety and
depressive disorders,21,22 identifying patients at high risk for relapse is of utmost impor-
tance. Especially in anxiety disorders, predictors of relapse are scarce.23 Our findings
thus implicate that OC-symptoms may be valuable specifiers in the context of both anx-
iety and depressive disorders, like, for example, anxiety symptoms in depression and
panic attacks across the full range of psychopathology.24,25 This knowledge could add
to models for profiling and staging of anxiety and depressive disorders that are current-
ly being developed in order to improve clinical prediction and to direct treatment.26-28
Our results showed that clinically relevant OC-symptoms do not predict worse out-
come independent of anxiety and depression severity. This could be explained by the
overlap of concepts between OCD on the one hand, and anxiety and depression on
the other hand; i.e. correlations between the YASR-OCS and questionnaires used to
assess severity of depression and anxiety showed a correlation of respectively 0.66
(between YASR-OCS and IDS) and 0.58 (between YASR-OCS and BAI). One might
therefore argue that OCD, anxiety and depression share symptom overlap and are less
distinctive disorders than suggested by DSM IV categorizations. Rather than consid-
ering OCD, anxiety disorders and depression as distinct disorders, which the original
71
Chapter 2.3
definition of comorbidity requires,29 this type of comorbidity might be regarded a cor-

relate for severity of the index disorder.

Several strengths and limitations should be taken into account. Strengths include the
large sample size, the well-implemented design of the NESDA study and the longitudi-
nal design. Three limitations should be acknowledged as well.
First, as OC-symptoms were first assessed at the second wave, of NESDA, we had
no other option than to use this as our baseline. This may have influenced results, as
attrition and exclusion was somewhat selective.
Second, since the diagnosis of OCD was not assessed in the NESDA study, we
resorted to clinically relevant OC-symptoms as assessed by the YASR-OCS. However,
since the cut-off score that was used to define clinically relevant OC-symptoms has
been found to adequately predict OCD, we are rather confident that the YASR-OCS
screener is a reliable indicator of OC symptoms.10 Moreover, even though not all pa-
tients may fulfill criteria of OCD, results showed that a substantial level of OC-symp-
toms in itself has a predictive value. In our opinion, impact of OCD is most likely even
larger than the results presented here.
Third, we used the term ‘persistence’ of symptoms to describe an unfavorable
course trajectory, defined as the presence of a disorder at the two year follow up. This
measure does not take into account the in-between trajectory and residual symptoms.
However, when these aspects would have been taken into account, course would be
even more unfavourable than the results presented here.
Finally, due to limited numbers with a first onset of an anxiety or depressive disor-
der, conclusions are preliminary.
Conclusion
Clinically relevant OC-symptoms (and probably OCD) are present in a substantial pro-
portion of patients with anxiety and depressive disorders and are associated with with
severity and unfavourable course trajectories of these disorders. Although it can con-
ceptually not be ruled out that clinically relevant OC-symptoms are a correlate of sever-
ity rather than a distinct disease entity, clinically relevant OC-symptoms might warrant
treatment and moreover, should alert clinicians for unfavourable outcomes in anxiety
and depressive disorders.
Acknowledgements
The infrastructure for the NESDA study is supported by participating universities and
mental health care organizations (VU University Medical Center, GGZinGeest, Arkin,
Leiden University Medical Center, GGZ Rivierduinen, University Medical Center Gron-
ingen, Lentis, GGZ Friesland, GGZ Drenthe, IQ Healthcare, Netherlands Institute for
Health Services Research (NIVEL) and Netherlands Institute of Mental Health and Ad-
diction (Trimbos).
72
Reference list
1. Kessler RC, McGonagle KA, Zhao S, et al. Lifetime and 12-month prevalence of
DSM-III-R psychiatric disorders in the United States. Results from the National Co-
morbidity Survey. Arch Gen Psychiatry. 1994;51:8-19.
2. Angst J and Vollrath M. The natural history of anxiety disorders. Acta Psychiatr
Scand. 1991;84:446-452.
3. Merikangas KE, Zhang H, Avenevoli S, et al. Longitudinal Trajectories of Depres-
Arch Gen Psych. 2003;60:993-1000.
4. Rush AJ, Zimmerman M, Wisniewski SR, et al. Comorbid psychiatric disorders
in depressed outpatients: Demographic and clinical features. J Affect Disord .
2005;87:43-55.
5. Fichter MM, Quadflieg N, Fischer UC, Kohlboeck G. Twenty-five-year course and
outcome in anxiety and depression in the Upper Bavarian Longitudinal Community
Study. Acta Psychiatr Scand. 2010;122:75-85.
6. Kessler RC and Frank RG. The impact of psychiatric disorders on work loss days.
Psychol Med. 1997;27 :861-873. 2.3
7. Fullana MA, Mataix-Cols D, Caspi A, et al. Obsessions and Compulsions in the
Community: Prevalence, Interference, Help-Seeking, Developmental Stability and
Co-Occurring Psychiatric Conditions. J Am Psych. 2009;166:329-336.
8. Penninx BW, Beekman AT, Smit JH, et al. The Netherlands Study of Depression
Res. 2008;17:121-140.
9. Achenbach, TM. Manual for the Young Adult Self Report and Young Adult Be-
havior Checklist. Department of Psychiatry, University of Vermont: Burlington VT,
1997.
10. Grootheest van DS, Cath DC, Beekman AT, Boomsma DI. Genetic and envi-
ronmental influences on obsessive-compulsive symptoms in adults: a popula-
tion-based twin-family study. Psychol Med. 2007;37:1635-1644.
11. Andrews G and Peters L. The psychometric properties of the Composite Inter-
national Diagnostic Interview. Soc Psychiatry Psychiatr Epidemiol. 1998;33:80–
88.
12. Wittchen HU. Reliability and validity studies of the WHO—Composite International
Diagnostic Interview (CIDI): a critical review. J Psychiatr Res. 1994;28:57–84.
13. Beck AT, Epstein N, Brown G, Steer RA. An inventory for measuring clinical anxi-
ety: psychometric properties. J Consult and Clin Psychol. 1988;56:893-897.
Inventory. J Anxiety Disord. 1992;6:55–61.
15. Rush AJ, Gullion CM, Basco MR, et al. The Inventory of Depressive Symptomatol-
ogy (IDS): psychometric properties. Psycholog Med. 1996;26:477-486.
16. Trivedi MH, Rush AJ, Ibrahim HM, et al. The Inventory of Depressive Symptom-
atology, Clinician Rating (IDS-C) and Self-Report (IDS-SR), and the Quick Inven-
tory of Depressive Symptomatology, Clinician Rating (QIDS-C) and Self-Report
(QIDS-SR) in public sector patients with mood disorders: a psychometric evalua-
tion. Psycholog Med. 2004;34:73-82.
73
Chapter 2.3
17. Bijl RV, Ravelli A, van Zessen G. Prevalence of psychiatric disorder in the general
population: results of The Netherlands Mental Health Survey and Incidence Study
(NEMESIS). Social Psychiatry Psychiatr Epidemiol. 1998;33:587-595.
18. García-Soriano G, Rufer M, Delsignore A, Weidt S. Factors associated with
non-treatment or delayed treatment seeking in OCD sufferers: a review of the liter-
ature. Psychiatry Res. 2014;220:1-10.
19. Klein Hofmeijer-Sevink M, van Oppen P, van Megen HJ, et al. Clinical relevance of
comorbidity in obsessive compulsive disorder: the Netherlands OCD Association
study. J Affect Disord. 2013;150:847-54.
20. Denys D, Tenney N, van Megen HJ, et al. Axis I and II comorbidity in a large sam-
ple of patients with Obsessive compulsive disorder. J Affect Disord. 2004:80;155-
162.
21. Bruce SE, Yonkers KA, Otto MW, et al. Influence of psychiatric comorbidity on
recovery and recurrence in generalized anxiety disorder, social phobia, and panic
disorder: a 12-year prospective study. Am J Psychiatry. 2005;162:1179–1187.
22. Hardeveld F, Spijker J, De Graaf R, et al. Prevalence and predictors of recur-
rence of major depressive disorder in the adult population. Acta Psychiatr Scand.
2010;122:184-91.
23. Scholten WD, Batelaan NM, van Balkom AJ, et al. Recurrence of anxiety disorders
and its predictors. Journal of Affective Disorders. 2013;147:180–185
orders, Fifth Edition. Washington, DC: American Psychiatric Association; 2014.
25. Batelaan NM, Rhebergen D, de Graaf R, et al. Panic Attacks as s Dimension of
Disorders and Level of Functioning. J Clin Psychiatry. 2012;73:1195-1202.
Sir Martin Roth. Psychother Psychosom. 2008;77:133-138.
27. Balkom van AJ, Oosterbaan DB, Batelaan N, et al. The characterization of anxiety
atrie. 2012;11:935-940.
28. Fava GA and Tossani E. Prodromal stage of major depression. Early Intervention
in Psychiatry. 2007;1: 9-18.
29. Feinstein AR. The pre-therapeutic classification of co-morbidity in chronic diseas-
es. J Chronic Dis. 1970;23:455–468.
74
2.3
75
76
Chapter 3
Expectancy bias and treatment outcome in patients with

panic disorder and agoraphobia
Puck Duits, Mieke Klein Hofmeijer-Sevink, Iris M. Engelhard, Joke M.P. Baas, J,
W.A. Ehrismann, Danielle C. Cath.
In review
Author Contributions: All of the authors were involved in the conception and design of this study and/or in
the analysis and interpretation of data for this manuscript. All of the authors drafted the manuscript and/or con-
tributed important intellectual content to its revision.
77
Chapter 3
Abstract
Background and objectives: Previous studies suggest that patients with panic disor-
der tend to overestimate the associations between fear-(ir)relevant stimuli and aversive
outcomes. This so-called expectancy bias is thought to play a role in the development
and treatment of anxiety disorders. The current study tested 1) whether patients with
panic disorder and agoraphobia (N = 71) show an expectancy bias to fear-irrelevant
and fear-relevant stimuli compared to healthy control subjects (N = 65), and 2) whether
expectancy bias before treatment predicts treatment outcome in a subset of these pa-
tients (n = 51).
Methods: In a computerized expectancy bias task, participants saw a series of pan-
ic-related and neutral words, and rated for each word the likelihood that it would be
followed by a loud, aversive sound.
Results: Results showed higher expectancy ratings to both panic-related and neutral
words in patients with panic disorder and agoraphobia compared to controls. Expec-
tancy ratings did not predict treatment outcome in the patient group.
Limitations: The present study only included self-reported expectancy ratings; future
research should add psychophysiological outcome measures to investigate both emo-
tional and cognitive aspects of fear. Furthermore, inclusion of a post-treatment expec-
tancy bias task is recommended to shed further light on the possibility that expectancy
bias might be attenuated by treatment itself.
Conclusions: Patients show higher expectancies of aversive outcome following both
fear-relevant and fear-irrelevant stimuli compared to controls, but this does not predict
treatment outcome.
78
Expectancy bias and treatment outcome in panic disorder
INTRODUCTION
Panic disorder is characterized by recurrent and unexpected panic attacks, defined as
sudden episodes of intense fear or discomfort that reaches a peak within 10 minutes,
during which at least 4 symptoms occur, such as breathing problems, sweating and
chest pain. Other characteristics include worries about the potential consequences
of the attacks and concerns about future panic attacks.1 Agoraphobia (i.e., situational
avoidance and fear) is present in a substantial minority of patients with panic disorder,
and is associated with increased disability, higher comorbidity rates and poorer treat-
ment outcome.2-4
Fear conditioning mechanisms entail a potential model to explain the development
of pathological anxiety and subsequent avoidance behavior through heightened fear
acquisition and impaired fear extinction processes. However, the tendency to over-
estimate the association between fear-(ir)relevant stimuli and aversive unconditioned
stimuli (US) may originate from pre-experimental expectancies, rather than from the
computational bias that occurs during a fear conditioning task.5 Elevated US expectan-
cies that are present preceding fear conditioning are typically referred to as the US ex-
pectancy bias or covariation bias,6 and may play an important role in the development
and maintenance of pathological anxiety.7-11 Previous experimental studies provided
evidence for the presence of US expectancy biases in panic-prone individuals12,13 and
in patients with a panic disorder14,15 compared to low-anxious, healthy controls. Further 3
evidence for the role of the US expectancy bias in the development of anxiety symp-
toms comes from a longitudinal study, in which enhanced US expectancy ratings pre-
dicted the development of posttraumatic stress symptoms in soldiers around 15 months
after their deployment to Iraq.16 Increased US expectancies were most pronounced
when subjects were exposed to fear-relevant (instead of fear-irrelevant) stimuli and
when US contingency was low.
In addition, enhanced US expectancy bias preceding therapy might also be an
important predictor of treatment outcome in patients with anxiety disorders. One pre-
vious study in subjects with spider phobia (N = 19) found significant predictive value of
post-treatment US expectancy ratings on long-term treatment success, suggesting that
the presence of higher US expectancy ratings predicts stronger return of fear two years
later.17 Furthermore, previous research suggested that US expectancy bias might be
attenuated by treatment, because enhanced US expectancy ratings that have been ob-
served in untreated patients with spider phobia are absent in patients who have been
treated for spider phobia.18,19
In the current study, an expectancy task was administered in patients with panic
disorder and agoraphobia (PD/A) and healthy controls. Patients with PD/A completed
the expectancy task before participating in a randomized controlled trial in which the
added value of D-cycloserine (versus placebo) to exposure therapy was studied. The
aim of the current study was to replicate and extend previous findings by examining 1)
whether patients with PD/A compared to healthy controls demonstrate higher US ex-
pectancy ratings to panic-related and neutral words before treatment, and 2) whether
US expectancy ratings measured before treatment would predict treatment outcome in
patients with PD/A. We hypothesized that 1) patients would show a stronger US expec-
tancy bias to fear-irrelevant and fear-relevant stimuli compared to healthy controls and
79
Chapter 3
that 2) higher US expectancy ratings before treatment would be associated with worse
treatment outcome in patients with PD/A.

Method
Participants
Seventy-one patients with PD/A were recruited through three mental health care or-
ganizations in the Netherlands: Altrecht Academic Anxiety Centre (Utrecht), GGZ
inGeest (Amsterdam), and GGZ Centraal (Ermelo). Patients with PD/A participated in
the current conditioning paradigm as part of a randomized controlled trial, in which the
effectiveness of D-cycloserine (DCS) in addition to exposure with response prevention
(ERP) was studied.20 Exclusion criteria for the current study were 1) dependence and/or
abuse of alcohol/drugs in the past three months, 2) current comorbid psychotic disor-
der, 3) current severe major depressive disorder, 4) current bipolar disorder, 5) mental
deficiency (verbal IQ < 80 as assessed with the Dutch Adult Reading test),21 and 6)
insufficient ability to speak or read Dutch. Diagnosis of PD/A and any comorbid diag-
noses were established with the Dutch version of the structured clinical interview for
DSM-IV Axis-I disorders (SCID-I).22,23 Thirty-eight patients (54 %) had no comorbid di-
agnosis, 15 patients (21 %) were diagnosed with one comorbid other anxiety disorder,
10 patients (14 %) with an additional mood disorder and 8 patients (11 %) were diag-
nosed with both a comorbid other anxiety disorder and a mood disorder. Thirty-two pa-
tients (45%) used at least one psychotropic medicine at time of participation, including
the use of serotonin reuptake inhibitors (N = 23), benzodiazepines (N = 13) and tricyclic
antidepressants (N = 2). Medication dosage was kept stable throughout the ERP.
Sixty-five healthy control subjects were recruited through advertisements (posters
and flyers) and via contacts of the researchers. Control subjects were matched with
the patient group on age, gender and level of education. Table 1 provides information
about the demographic and clinical characteristics of the patient and control groups.
Absence of a lifetime DSM-IV Axis I disorder in control subjects was confirmed by us-
ing the Mini International Neuropsychiatric Interview.24,25 None of the control subjects
used psychotropic medication.
This study was approved by the Medical Research Ethics Committee of the Univer-
sity Medical Centre Utrecht. Subjects were first informed about the study, both orally
and by written information, and then provided written informed consent.
Procedure
At baseline, prior to the first treatment session, the US expectancy bias task was ad-
ministered to patients with PD/A. The task developed by Engelhard16 and based on ex-
periments 2 and 4 by Davey6 was adapted for the current study, i.e., the deployment-re-
lated images used by Engelhard were replaced by panic-related and neutral words for
this study. Participants were seated in a quiet room and completed the US expectancy
bias task on a laptop. To increase the probability that participants would answer truth-
fully during the task, they were told that the purpose of the study was to collect physi-
ological responses to various words, and to link these with the participants’ subjective
ratings (cf. Engelhard).16 Subsequently, two fake electrodes were attached to two fin-
gertips, and the US expectancy task started. Verbal and written instructions were given
80
Table 1
Demographic and clinical characteristics of the participants.
Patient group (N = 71) Control group (N = 65) Significance of group

differences a
N % N %
Male gender 28 39.4 31 47.7 p = .332
Mean SD Mean SD
Age 34.8 10.2 37.9 13.6 p = .145
ACQ 2.19 .56 1.13 .41 p < .001
BSQ 2.64 .66 1.28 .19 p < .001
MI when alone 2.92 .97 1.23 .34 p < .001
MI when accompanied 2.18 .71 1.10 .21 p < .001
a
Two-tailed
ACQ = Agoraphobic Cognition Questionnaire; BSQ = Body Sensations Questionnaire;26
MI = the Mobility Inventory 27
to inform participants that words would be presented on the screen during the task,
and that some of these words might be followed by the aversive, loud sound. Through
headphones, participants then received a single presentation of a loud white noise (95
dB, 500 ms) that served as US. Participants were asked to indicate their expectancy of
the US during each trial on an online Visual Analogue Scale (VAS; 0 = certain no noise,
100 = certain noise), to index the extent to which they expected the aversive sound to
occur. VAS scales were continuously displayed during the task, and participants were
asked to rate their US expectancy at least once per presented word (cf. Engelhard).16
US expectancy ratings are commonly used in patients with panic disorder28-30 and are
regarded as valid measures to study cognitive processes in fear conditioning.31 During
the experiment, panic-related and neutral words were presented on the laptop screen
in fixed random order. The task consisted of three panic words: “panic”, “fear”, “anxiety”
and three neutral words “butter”, “carpet”, “sidewalk”; panic-related and neutral words
were matched on word length and familiarity in the Dutch language. The panic-related
words have been rated as more unpleasant than neutral words by 145 students in an
earlier study.32 Each word was presented four times for 4-5 seconds, with inter trial in-
tervals (a black screen) varying between 5-9 seconds. Words from the same category
(panic-related or neutral) were never presented more than twice consecutively. Halfway
through the experiment (i.e. after 6 words of both categories had been presented), one
single US reinforcement was given after presentation of a neutral word to reinstate the
US expectancy bias.6,16
81
Chapter 3
Treatment outcome
This study was part of a multi-center randomized controlled trial to examine the add-
ed value of DCS administration in patient with PD/A. Of 71 patients with PD/A who
completed the US expectancy bias task, 51 patients subsequently enrolled in the
multicenter double-blind placebo-controlled trial. The other 20 patients were either not
willing to participate in the randomized controlled trial, or were excluded for participa-
tion. For those 51 patients who did participate in both the US expectancy task and the
randomized controlled trial, the average mean time between the US expectancy bias
task and start of the treatment trial was 10 days (SD = 35). Within the treatment trial,
patients were randomized to receive either placebo (n = 19) or fixed dosages of 125
mg D-cycloserine during ERP, given directly preceding (n = 16) or directly post (n = 16)
session 2-7 of twelve 90 minute sessions ERP.
The treatment outcome measures were the Mobility Inventory27 and the Panic
Disorder Severity Scale.33 The MI measures self-reported avoidance, and contains
two subscale scores: 1) mobility when a patient is alone and 2) when the person is
accompanied by a trusted companion.34 The PDSS encompasses an interview to rate
the severity of panic symptoms in patients with established diagnoses. Both the MI
and PDSS were assessed at baseline (pre-treatment), after session 3, 7 and 12 (post
treatment), and at 3 and 6 months follow up. For the current analyses, only baseline
and post-treatment assessments were used. Treatment outcome was defined as the
percentage of improvement = ((pre-treatment score – post-treatment score) * 100) /
pre-treatment score. On group level, patients with PD/A demonstrated an average im-
provement of 27 % (SD = 22) on MI scores and 68 % (SD = 37) on the PDSS scores,
indicating that at least half of the included patients could be classified as treatment re-
sponders, since these patients showed ≥25 % improvement.35-37 Treatment outcome of
DCS enhancement suggested no additional effect of DCS administration compared to
placebo, and results are elaborately reported separately.20
Statistical Analyses
Using IBM SPSS (version 22), repeated measures ANOVA was conducted on the base-
line measurements to investigate whether patients with PD/A showed higher US expec-
tancy ratings to panic-related and neutral words compared to controls. Group (patients
versus controls) was the between-subjects factor, and stimulus type (panic-related ver-
sus neutral words) and expectancy trial number (12 data points) were included as with-
in-subjects factors. Because participants were instructed to rate their expectancy at
least once per presented word, data reduction was applied when there was more than
one rating on one trial. In accordance with a previous similar study, the highest rating
from each trial was selected for the analyses.16 When the assumption of sphericity was
not met, Greenhouse-Geisser correction (ε < .75) or Huynh-Feldt correction (ε > .75)
were applied to the F-ratio and degrees of freedom.
Linear regression analyses (enter method) were carried out to examine whether US
expectancy ratings predicted treatment outcome (MI and PDSS, as outlined above) in
patients with PD/A, while controlling for the possible influence of DCS. US expectancy
ratings were included separately for stimulus type (panic-related or neutral words) and
stage (before or after reinstatement). Data from ten patients could not be included in
82
the regression analyses wherein MI served as outcome variable, because these pa-
tients did not complete the MI at both pre- and post-treatment measurement.
Both US expectancy ratings and the use of DCS versus placebo were included as
predictors in each regression model. Furthermore, interaction between DCS and US
expectancy on treatment outcome was investigated, hypothesizing that DCS works
via extinction learning and that patients with strong US expectancy biases might profit
more from the enhancing effect of DCS on treatment outcome, as these patients have
more room for improvement when down regulating their heightened US expectancies.
Finally, since a substantial proportion of patients (N = 13) used benzodiazepines at
time of testing, which might have influenced US expectancy ratings in the current para-
digm, analyses were repeated while excluding the patients who used benzodiazepines
during participation. However, exclusion of these patients did not influence the results
(data not shown) and therefore results are reported including the whole group.
Figure 1
Figure 1. US expectancy ratings regarding panic-related words (left graph) and
US expectancy ratings regarding panic-related words (left graph) and neutral words (right graph)
neutral words (right
in patients graph)
with PD/A in patients
and control with PD/A and control subjects
subjects.
RESULTS
Figure 1 displays baseline US expectancy ratings in patients with PD/A and control sub-
jects; ratings are shown separately for panic-related words (left graph) and neutral words
(right graph). Results demonstrated a significant main effect for stimulus type, F(1,132) =
129.9, p < .001, ɳp2 = .50, which reflects elevated US expectancy ratings to panic-related
words compared to neutral words across patients and controls and across all expec-
tancy trials. A significant main effect of trial indicated that the US expectancy task was
sensitive to changes over time, F(7.9, 1040.1) = 4.6, p < .001, ɳp2 = .03. The contrast for
within subjects factors reflected a decline over time in some trials (trial 7-8, 8-9, and 10-
11), while other trials showed an increase over time (trial 4-5, 9-10, and 11-12). Further-
more, a significant interaction between stimulus type and trial was found across groups
between trial 6 and 7, reflecting an increase in US expectancy ratings towards neutral
words and a decrease in US expectancy towards panic-related words as a consequence
of the single US presentation preceding trial 7, F(1, 132) = 8.6, p = .004, ɳp2 = .06.
83
Chapter 3
A significant main effect of group was found, F(1,132) = 24.0, p < .001, ɳp2 = .15, indi-
cating overall increased expectancy ratings in patients with PD/A compared to controls
(see figure 1). The group by trial interaction, F(7.9, 1040.0) = .94, p = .50, ɳp2 = .01, and
group by stimulus type by trial interaction, F(11, 1452) = 1.13, p = .33, ɳp2 = .01, were not
significant.
Regression analyses demonstrated no significant predictive value of US expectan-
cy ratings on treatment outcome. Furthermore, neither administration of DCS, nor the
interaction between US expectancy ratings and DCS were associated with treatment
outcome. To sum up, none of the predictor variables accounted significantly for the
variability in treatment outcome measured with the two subscales of the MI (i.e., mo-
bility when a patient is alone and when the patient is accompanied by a trusted com-
panion) or the PDSS. Table 2 provides an overview of the model fits, displayed for each
treatment outcome per stimulus type (panic-related versus neutral words) and stage
(before versus after reinstatement).
Table 2
Results of regression analyses predicting treatment outcome in patients with PD/A
Model Fits MI when alone MI when accompanied PDSS
R2 df F p R2 df F p R2 df F p
Panic word - .09 3,40 1.2 .315 .04 3,39 .5 .692 .08 3,49 1.3 .281
stage1
Panic word - .07 3,40 .9 .431 .06 3,39 .7 .541 .06 3,49 1.0 .421
stage 2
Neutral word - .10 3,40 1.4 .263 .07 3,39 .9 .447 .10 3,49 1.8 .166
stage 1
Neutral word - .17 3,40 2.5 .072 .15 3,39 2.1 .121 .07 3,49 1.1 .370
stage 2
MI = Mobility Inventory; PDSS = Panic Disorder Severity Scale; Stage 1 = before reinstatement;
Stage 2 = after reinstatement.
DISCUSSION
In this study, we examined two groups of 71 patients with a panic disorder and agora-
phobia and 65 matched control subjects to investigate 1) differences in US expectancy
ratings, and 2) the predictive value of US expectancy on treatment outcome, while
controlling for administration of DCS. Results demonstrated higher US expectancy rat-
ings to both panic-related and neutral words in patients with PD/A compared to healthy
control subjects. The observed US expectancy bias in patients with PD/A persisted
throughout the paradigm, even though implicitly, information on US expectancy was
84
disconfirmed (i.e. the aversive event was never paired with a neutral or panic-related
word). These overall higher US expectancy ratings in patients with PD/A compared to
controls are in line with previous findings12-15 and suggest that patients with PD/A tend
to (persistently) overestimate stimuli-US contingencies, which has been associated
with an increased risk of developing anxiety symptoms16 and stronger relapse of anxi-
ety symptoms at two years after treatment.17
This is the first study in which the predictive value of baseline US expectancy
bias on treatment outcome was investigated in patients with in PD/A. Based on one
earlier study in spider phobics, we hypothesized that higher US expectancy ratings
before treatment would be associated with worse treatment outcome in patients with
PD/A. However, results demonstrated no significant predictive value of baseline US
expectancy ratings (and DCS) on treatment outcome. This suggests that pre-treatment
expectancy bias is not related to treatment outcome, and we should focus on other pre-
dictors for treatment outcome instead. An example of such a potential predictor is the
extinction of fear, which might serve as an underlying factor of the development of anx-
iety disorders,16,38-39 and exposure therapy.40-42 Support for the predictive value of fear
extinction on treatment outcome was found in recent studies conducted in patients with
panic disorder and agoraphobia.43-44Treatment non-responders displayed enhanced
activation of threat-related brain systems in response to safety cues during extinction
when compared to responders.45 Another explanation for the current findings may be 3
that the expectancy bias task was not sensitive enough to examine the relationship be-
tween expectancy ratings and treatment outcome. The study was limited by exclusively
using subjective US expectancy ratings, which reflect more cognitive based processes
that might be susceptible to experimental demands, such as participants trying to de-
termine the purpose of the experiment.31 Physiological outcome measurements (such
as startle potentiation) might on the other hand represent a different aspect of anxiety;
it has been suggested that physiological measurements reflect the emotional instead
of cognitive component of anxiety.46 Addition of physiological outcome measurements
is recommended for future studies to cover both emotional and cognitive aspects of
anxiety.
The discrepancy between previous findings by de Jong and colleagues,17 who did
find a stronger return of fear after two years in spider phobics with high US expectancy
ratings before treatment, and current findings may be explained by methodological
differences. First, presentation of fear-(ir)relevant stimuli was partly reinforced by the
US in the previous study, while no panic-related or neutral words were reinforced in the
current study (instead, one single US was presented halfway the experiment, which
was not paired with a word; cf. Engelhard et al., 2009;16 Davey, 1992, exp 2 and 4).6 As
a consequence, the study of de Jong and colleagues may have been more related to
fear learning, while the current study focused more specifically on US expectancy bias
that is present prior to fear conditioning. Second, the previous study assessed expec-
tancy bias after treatment, while the current study examined expectancy bias before
treatment. For future studies, we recommend the use of both pre- and post-treatment
measurements within one experiment to extent our knowledge about the course of the
US expectancy bias as a function of treatment outcome. Other differences between the
previous versus current study that may explain the discrepancy in findings: differences
85
Chapter 3
in populations (patients with spider phobia versus patients with PD/A), sample sizes (19
patients versus 51 patients), duration of treatment (one 2.5-hr exposure session versus
twelve 1.5-hr exposure sessions) and timing of treatment outcome measurement (two
years after treatment versus measurement after twelve therapy sessions).
To conclude, higher US expectancy ratings to both panic-related and neutral words
were found in patients with PD/A compared to control subject, but these increased rat-
ings did not predict treatment outcome in patients with PD/A. Future research should
add both psychophysiological and subjective outcome measurements and we recom-
mend the use of a pre- and post-treatment US expectancy task to shed further light on
the possibility that the expectancy bias is down-regulated as a result of treatment. By
extending our knowledge about the relationship between expectancy bias and treat-
ment outcome, tailored treatment may be developed for those patients who overesti-
mate stimuli-US contingencies.
Acknowledgements
We thank the participants for their involvement, and the participating centers (Altrecht
Academic Anxiety Centre, GGZ inGeest and GGZ Centraal) for their cooperation.
86
Reference list
1. American Psychiatric Association. The diagnostic and statistical manual of mental
disorders: DSM 5. Bookpoint US; 2013.
2001;110:585-599.
3. Slaap BR, Den Boer JA. The prediction of nonresponse to pharmacotherapy in
panic disorder: A review. Depress Anxiety 2001;14:112-122.
4. Kessler RC, Chiu WT, Jin R et al. The epidemiology of panic attacks, panic dis-
order, and agoraphobia in the national comorbidity survey replication. Arch Gen
Psychiatry 2006;63:415-424.
5. Davey GC. A conditioning model of phobias. Phobias: A handbook of theory, re-
search and treatment 1997:301–322.
6. Davey GC. An expectancy model of laboratory preparedness effects. J Exp Psy-
chol: Gen 1992;121:24-40.
7. Öhman A, Mineka S. Fears, phobias, and preparedness: Toward an evolved mod-
ule of fear and fear learning. Psychol Rev 2001;108:483-522.
8. McNally RJ. Psychological approaches to panic disorder: A review. Psychol Bull
1990;108:403-419.
9. Tomarken AJ, Mineka S, Cook M. Fear-relevant selective associations and covari- 3
ation bias. J Abnorm Psychol 1989;98:381-394.
10. Davey GC. Cognitive mechanisms in fear acquisition and maintenance. In M. G.
Craske, D. Hermans, & D. Van Steenwegen (Eds.), Fear and learning (pp. 99–116).
Washington, DC: American Psychological Association; 2006.
11. Vroling MS, De Jong PJ. Belief bias and the extinction of induced fear. Cognition &
emotion 2013;27:1405-1420.
12. Pauli P, Montoya P, Martz G. Covariation bias in panic-prone individuals. J Abnorm
Psychol 1996;105:658-662.
13. Pauli P, Montoya P, Martz G. On-line and a posteriori covariation estimates in pan-
ic-prone individuals: Effects of a high contingency of shocks following fear-irrele-
vant stimuli. Cognitive Therapy and Research 2001;25:23-36.
14. Amrhein C, Pauli P, Dengler W, Wiedemann G. Covariation bias and its physiologi-
cal correlates in panic disorder patients. J Anxiety Disord 2005;19:177-191.
disorder. J Anxiety Disord 2001;15:401-412.
16. Engelhard IM, de Jong PJ, van den Hout MA, van Overveld M. Expectancy bias
and the persistence of posttraumatic stress. Behav Res Ther.2009;47:887-892.
17. De Jong PJ, Van Den Hout, Marcel A, Merckelbach H. Covariation bias and the
return of fear. Behav Res Ther 1995;33:211-213.
18. de Jong PJ, Merckelbach H, Arntz A, Nijmam H. Covariation detection in treated
and untreated spider phobics. J Abnorm Psychol 1992;101:724-727.
19. de Jong PJ, Merckelbach H, Arntz A. Covariation bias in phobic women: The rela-
tionship between a priori expectancy, on-line expectancy, autonomic responding,
and a posteriori contingency judgment. J Abnorm Psychol. 1995;104(1):55-62.
20. Klein Hofmeijer-Sevink M, Duits P, Rijkeboer M.M et al. D-cycloserine addition in
87
Chapter 3
exposure therapy for patients with panic disorder with agoraphobia; a randomized
controlled trial. Unpublished results.
21. Schmand B, Bakker D, Saan R, Louman J. De nederlandse leestest voor volwas-
senen: Een maat voor het premorbide intelligentieniveau./the dutch adult reading
test: A measure of premorbid intelligence. Tijdschrift Voor Gerontologie En Geri-
atrie 1991;22:15-19.
23. Groenestijn M, Akkerhuis GW, Kupka RW et al. Gestructureerd klinisch interview
voor de vaststelling van DSM-IV as I stoornissen (SCID I). Lisse: Swets test; 1997.
24. Lecrubier Y, Sheehan D, Weiller E et al. The mini international neuropsychiatric
interview (MINI). A short diagnostic structured interview: Reliability and validity ac-
cording to the CIDI. European psychiatry 1997;12:224-231.
25. Sheehan D, Lecrubier Y, Harnett Sheehan K et al. The validity of the mini interna-
tional neuropsychiatric interview (MINI) according to the SCID-P and its reliability.
European Psychiatry 1997;12:232-241.
26. Chambless DL, Caputo GC, Bright P et al. Assessment of” fear of fear” in agora-
phobics: The body sensations questionnaire and the agoraphobic cognitions ques-
tionnaire. Journal of Consulting and Clinical Psychology 1984;52:1090-1097.
27. Chambless DL, Caputo GC, Jasin SE et al. The mobility inventory for agoraphobia.
Behav Res Ther 1985;23:35-44.
28. Grillon C, Lissek S, Rabin S et al. Increased anxiety during anticipation of unpre-
dictable but not predictable aversive stimuli as a psychophysiologic marker of pan-
ic disorder. Am J Psychiatry 2008;165:898-904.
29. Lissek S, Rabin SJ, McDowell DJ et al. Impaired discriminative fear-conditioning
resulting from elevated fear responding to learned safety cues among individuals
with panic disorder. Behav Res Ther 2009;47:111-118.
30. Lissek S, Rabin S, Heller RE et al. Overgeneralization of conditioned fear as a
pathogenic marker of panic disorder. Am J Psychiatry 2010;167:47-55.
31. Boddez Y, Baeyens F, Luyten L et al. Rating data are underrated: Validity of US
expectancy in human fear conditioning. J Behav Ther Exp Psychiatry 2013;44:201-
206.
32. Hermans D, De Houwer J. Affective and subjective familiarity ratings of 740 dutch
words. Psychologica Belgica 1994;35:115-139.
33. Shear MK, Brown TA, Barlow DH et al. Multicenter collaborative panic disorder se-
verity scale. Am J Psychiatry 1997;154:1571-1575.
34. Chambless DL, SharplessBA, Rodriguez D et al. Psychometric properties of the
mobility inventory for agoraphobia: Convergent, discriminant, and criterion-related
validity. Behavior Therapy 2011;42:689-699.
35. Buchsbaum MS, Hollander E, Pallanti S et al. Positron emission tomography imag-
ing of risperidone augmentation in serotonin reuptake inhibitor-refractory patients.
Neuropsychobiology 2006;53:157-168.
36. Saxena S, Brody AL, Maidment KM et al. Localized orbitofrontal and subcortical
metabolic changes and predictors of response to paroxetine treatment in obses-
88
sive-compulsive disorder. Neuropsychopharmacology 1999;21:683-693.

37. Ho Pian KL, van Megen HJ, Ramsey NF et al. Decreased thalamic blood flow in
obsessive-compulsive disorder patients responding to fluvoxamine. Psychiatry Re-
search: Neuroimaging 2005;138:89-97.
38. Guthrie RM, Bryant RA. Extinction learning before trauma and subsequent post-
traumatic stress. Psychosom Med 2006;68:307-311.
39. Lommen MJ, Engelhard IM, Sijbrandij M et al. Pre-trauma individual differences in
extinction learning predict posttraumatic stress. Behaviour Research and Therapy
2013;51:63-67.
40. Hofmann SG. Cognitive processes during fear acquisition and extinction in an-
imals and humans: Implications for exposure therapy of anxiety disorders. Clin
Psychol Rev 2008;28:199-210.
41. Massad PM, Hulsey TL. Exposure therapy renewed. Journal of Psychotherapy In-
tegration 2006;16:417-428.
42. Myers KM, Davis M. Chapter 2.3 extinction of fear: From animal studies to clinical
interventions. In: Handbook of behavioral neuroscience. Vol Volume 17 (pag 49-
62) Elsevier; 2008.
43. Kircher T, Arolt V, Jansen A et al. Effect of cognitive-behavioral therapy on neural
correlates of fear conditioning in panic disorder. Biol Psychiatry 2012;73:93-101.
44. Lueken U, Straube B, Konrad C et al. Neural substrates of treatment response to 3
cognitive-behavioral therapy in panic disorder with agoraphobia. Am J Psychiatry
2013;170:1345-1355.
45. Hahn T, Kircher T, Straube B et al. Predicting treatment response to cognitive
behavioral therapy in panic disorder with agoraphobia by integrating local neural
information. JAMA Psychiatry 2014;72:68-74.
46. Sevenster D, Beckers T, Kindt M. Retrieval per se is not sufficient to trigger recon-
solidation of human fear memory. Neurobiol Learn Mem 2012;97:338-345.
89
90
Chapter 4
D-cycloserine does not enhance exposure therapy in

patients with panic disorder with agoraphobia: a double-blind,
randomized controlled trial
Mieke Klein Hofmeijer-Sevink, Puck Duits, Marleen M. Rijkeboer, Adriaan W.

Hoogendoorn, Harold J. van Megen, Nienke C. Vulink, Damiaan A. Denys, Marcel
A. van den Hout, Anton J. van Balkom, Danielle C. Cath.
Submitted for publication
Author contributions: All of the authors were involved in the conception and design of this study and/
or in the analysis and interpretation of data for this manuscript. All of the authors drafted the manus-
cript and/or contributed important intellectual content to its revision.
91
Chapter 4
Abstract
Objective: D-cycloserine (DCS) is a partial NMDA receptor agonist that by enhancing
extinction learning potentially augments response to exposure therapy in anxiety dis-
orders. This randomized, double-blinded, placebo-controlled augmentation trial exam-
ined i) the effectiveness of adding 125 mg DCS to exposure therapy (before or directly
after) in patients with panic disorder with agoraphobia and ii) the effectiveness of DCS
augmentation prior to exposure, as compared to DCS augmentation after exposure.
Method: Patients with a primary diagnosis of panic disorder with agoraphobia (n=57)
were allocated to one of three medication conditions (placebo, DCS pre and DCS post
exposure) as an addition to six exposure sessions within a twelve-session exposure
protocol. The primary outcome measure was the mean score on the “alone” subscale
of the Mobility Inventory (MI), the secondary outcome measure was the mean score on
the “accompanied” subscale of the MI, Beck Anxiety Inventory (BAI), the Panic Disor-
der Severity Scale (PDSS) and the Beck Depression Inventory-II (BDI-II).
Results: No differences in treatment outcome were found between DCS and placebo,
either administered pre or post exposure therapy. Ancillary analyses in specific patient
subgroups (responders versus non-responders, early versus late responders, severely
versus mildly affected patients) did not reveal any between-group DCS versus placebo
differences. Neither DCS pre nor DCS post exposure showed additional improvement
of exposure therapy compared to placebo.
Conclusions: Slightly in contrast with previous research and one meta-analysis, this
study showed no enhancing effects of DCS augmentation to exposure therapy in panic
disorder with agoraphobia. Although we cannot rule out a small effect of DCS which we
might have failed to pick up due to the relatively small sample size of this study, DCS
does not seem to hold much promise in the clinical treatment of patients with panic dis-
order and agoraphobia.
92
D-cycloserine in panic disorder
Introduction
Panic disorder is among the most prevalent anxiety disorders, with a twelve-month
prevalence rate estimated at around 2.7%.1 Current treatment involves cognitive be-
havioral therapy (CBT), medication (predominantly SSRIs and benzodiazepines) or the
combination, with combined treatment demonstrating an advantage over monothera-
py.2-4 Although effect sizes of treatment are moderate to large,5,6 relapse is common7,8
and there is still substantial room for improvement, particularly in patients at the more
severe and chronic end of the spectrum.9 Therefore, current research focuses on alter-
native treatment-enhancing strategies.
One of the pharmacological enhancement targets is the N-methyl-D-aspartate
(NMDA) receptor in the amygdala that is involved in the acquisition, consolidation, re-
consolidation and extinction of fear memory.10 Animal research has indicated that d-cy-
closerine (DCS), a partial NMDA agonist, can enhance extinction of fear memory.11-13
Since extinction learning is considered to be the core mechanism of exposure thera-
py,14,15 these results suggest that DCS can provide benefits for clinical treatment.
So far, thirteen clinical studies have investigated the potential enhancing effect of
DCS as an addition to CBT in anxiety disorders.16-28 Meta-analyses of these studies
found small to moderate observed effect sizes (Cohen’s d between 0.34-0.42) and
suggested that DCS augmentation is most effective when administered for a limited
number of times and at low dosages.29-31 None of the clinical studies has directly com-
pared DCS augmentation pre exposure therapy versus post exposure therapy. This is
of clinical importance since post exposure administration of DCS offers the advantage
of augmenting successful sessions only. There is evidence from animal studies that 4
DCS also enhances fear conditioning processes, and therefore carries the risk of erro-
neously enhancing fear memories in unsuccessful exposure sessions as well.32 Finally,
previous animal studies found DCS augmentation post extinction learning experiments
to be equally effective as directly preceding them.11,33 To our knowledge, direct compar-
ative studies in humans are lacking.
We aimed at replicating and extending previous findings of DCS augmentation in
exposure therapy in a large group of patients with panic disorder with agoraphobia.
The first aim of this study was to compare the addition of DCS to exposure therapy with
placebo. We hypothesized that DCS would enhance and or speed up the effect of ex-
posure therapy. Furthermore, aiming to replicate previous studies that found additional
benefit of DCS in specific subgroups ,23,24,34,35 we examined the added effect of DCS in
these in subgroups (i.e. responders versus non-responders, early versus late respond-
ers, severely versus mildly affected patients). The second aim of the study was to
compare DCS administration 30 minutes before exposure sessions (DCS pre) to DCS
administration directly after exposure sessions (DCS post). Our aim was to explore any
differences between DCS pre and post exposure.

93
Chapter 4
Figure 1
Patients flow chart.
Figure 1. Patients flow chart
Recrutment and enrollment
196 invited for baseline N= 41 (21%) not included:

interview - no consent (n=23),
- no panic disorder with agoraphobia (n=13),
- pregnancy wish (n=2),
- inability to adequately read or speak Dutch (n=3)
155 baseline interview N= 98 (63%) not included:

- no consent (n=40),
- no SCID-I panic disorder with agoraphobia (n=38),
- IQ<80 (n=6),
- use of day time benzodiazepines (n=7),
- CBT prior to study (n=2),
57 included and randomized - renal failure (n=1),
- inability to adequately read or speak Dutch (n=1),
- allergic reaction to penicilline (n=1),
- no reason (n=2).
Allocation
DCS prior to an exposure DCS after an exposure Placebo (n=19)

session (n=19) session (n=19)
Intervention Assessments:
- After session 3 (mid-study
1 introduction session medication period)
6 sessions with study medication - After session 7 (post –
5 sessions without study medication study medication period)
- After session 12 (post
exposure therapy period)
- Completers (n=15, 79%) - Completers (n=16, 84%) - Completers (n=14, 73%)

- Drop outs (n=3, 16%) - Drop outs (n=2, 11%) - Drop outs (n=3, 16%)
- Early remission (n=1, 5%) - Early remission (n=1, 5%) - Early remission (n=2,11%)
- Lost to follow up after 3 - Lost to follow up after 3 - Lost to follow up after 3
months (n=2, 11%) months (n=2, 5%) months (n=1, 5%)
- Lost to follow up after 6 - Lost to follow up after 6 - Lost to follow up after 6
months (n=1, 5%) months (n=1, 5%) months (n=1, 5%)
94
Method
Patients
Patients were recruited between October 2010 and October 2013 at the outpatient clin-
ics in three participating Dutch mental health care institutions. This study was approved
by the local ethics committee, registered at www.trialregister.nl (identifier: 6577) and
all subjects gave their written informed consent to participate. Adult patients (n=196)
who had been referred for a panic disorder with agoraphobia were invited to a baseline
structured interview. The Dutch version of the Structured Clinical Interview for DSM-IV
Axis I diagnoses (SCID-I)36,37 was used to determine in- and exclusion criteria for the
study. Co-morbidities including severe major depressive disorder (score >29 assessed
with the Beck Depression Inventory II (BDI-II),38 bipolar disorder, current psychotic
disorder and dependence and/or abuse of alcohol/drugs during the past three months
served as exclusion criteria. Other exclusion criteria were: intellectual disability (Verbal
IQ <80 as assessed with the Dutch Reading Test for Adults);39 an inability to adequate-
ly read or speak Dutch; a history of neurological disease, renal or liver abnormalities;
pregnancy or lactation; a history of severe adverse reactions to penicillin; and an
unsuccessful evidence-based behavioral therapy for panic disorder in the preceding
twelve months. Finally, current daily daytime use of benzodiazepines was an exclusion
criterion. SSRIs were allowed as long as the dosage was kept stable from at least three
months prior to the start of the study until the end of the study.
Twenty-three patients refused to be interviewed and could therefore not be invited
to participate and eighteen patients were lost before intake (Figure 1). Hence, 155 pa-
tients were interviewed. Ninety-eight patients were excluded because a) they did not 4
fulfill a SCID-I diagnosis of panic disorder with agoraphobia (n=38); b) they refused to
participate in the study (n=40); or c) they met one of the other exclusion criteria (n=20,
for details, see Figure 1). Fifty-seven patients were randomized to either DCS prior to
an exposure session (n=19), DCS after an exposure session (n=19) or placebo (n=19).
Psychometric assessments
Patients were assessed before the beginning of therapy (baseline), after session three
(mid-study medication period), session seven (post-study medication period), session
twelve (post-exposure therapy) and at a three- and six-month follow-up. The primary
outcome measure was the mean score on the “alone” subscale of the Mobility Invento-
ry (MI).40 This self-report questionnaire has good reliability and validity41 and contains
27 items (1-5 point scale per item) that assess agoraphobic avoidance behavior; situ-
ations are rated on amount of avoidance “when alone” and “when accompanied by a
significant other’. As secondary outcome measures we assessed mean scores of the
MI “accompanied” subscale, of the Beck Anxiety Inventory (BAI),42 of the Panic Dis-
order Severity Scale (PDSS),43 and the Beck Depression Inventory II (BDI-II).38 These
questionnaires are proven to be valid and reliable.44-47
Finally, we used the Fawcett side effects checklist48 to monitor possible side effects
as a result of DCS administration at all assessments.
Exposure therapy
Patients underwent twelve 90-minute individual sessions of exposure therapy. The
95
Chapter 4
standardized treatment protocol was based on evidence-based CBT manuals for pan-
ic disorder with agoraphobia.49,50 The first session provided patient education and an
idiosyncratic fear hierarchy was constructed. The next six sessions (which were aug-
mented with DCS) all included interoceptive or in-vivo exposure, as well as elements
of education and cognitive restructuring. Sessions eight to twelve (without study med-
ication) included in-vivo exposure, cognitive restructuring, and the construction of a
relapse-prevention plan.
Therapists were licensed psychologists and psychiatrists. They were well trained
in CBT and in the study protocol, and they took part in monthly supervision sessions
during the study. All sessions were audio taped (unless this was impossible due to
exposure-in-vivo) and five percent of the tapes were randomly checked to guarantee
treatment integrity.
Design, dosing, randomization and monitoring of study medication

This study entailed a randomized, double-blinded, placebo-controlled, fixed-dose mul-
ticenter trial, with DCS as an adjunct to six out of twelve weekly exposure sessions and
with three treatment arms. Patients in condition one (DCS pre) received 125 mg DCS
half an hour prior to the first six exposure sessions, and placebo directly after the first
six exposure sessions. Patients in condition two (DCS post) received placebo half an
hour before the first six exposure sessions and DCS directly after the first six exposure
sessions. Patients in condition three (placebo) received placebo both half an hour prior
to and directly after the first six exposure sessions.
Sample size calculations were based on two groups (DCS versus placebo), on a
0.05 significance level (two-tailed) and a power of 80%. Calculations were based on
the only available enhancement study using DCS prior to exposure therapy in panic
disorder at the time of our sample size calculation,22 which yielded an effect size of 1.1
on DCS enhancement. To find an effect size (Cohen’s d) of 1, a per-protocol sample
size would require 20 patients per condition, with a two-sided alpha set at 0.05, and a
beta of 0.20.
The medication used in the study was prepared by our pharmacy department,
which also carried out the randomization procedure. The randomization sequence was
guarded by the pharmacy until the last follow-up data had been collected. Thus, both
researchers and patients were blind to the allocation sequence. The study medication
was distributed by the therapist.
Statistical analysis
Data were analyzed using SPSS version 19 (SPSS Inc., Chicago, Illinois). Be-
tween-group differences in baseline descriptives were compared using 2-tailed χ2
statistics for categorical variables and 1-way Analysis of Variance (ANOVA) statistics
for continuous variables. Linear mixed model (LMM) procedures for repeated measure
analyses were performed, allowing all patients to be included in the analyses regard-
less of missing data.
To study differences in effect between DCS and placebo, a LMM procedure was
carried out with two groups (DCS versus placebo) and six different times (baseline,
mid-study medication period, post study medication period, post exposure therapy, at
96
three-months and at six-months follow-up) as the fixed effects and the time x group
interaction terms. Between group effect sizes are based on predicted means and stan-
dardized using pooled baseline SDs according to the PPC2 method.51 Effect sizes are
reported as Cohen’s d.
To study whether DCS facilitates treatment effect in specific patient groups, we
examined the following patient subgroups: i) responders versus non-responders, with
response defined as ≥ 25% symptom reduction on the MI “alone” score, measured
directly post exposure therapy (after session twelve); ii) early responders versus late
responders, with early response defined as ≥25% symptom reduction on the MI “alone”
subscale directly post session seven versus later or no response; iii) severely versus
mildly affected patients at baseline (severely affected being defined as showing a
baseline MI “alone” score above the median of 2.9.
LMM was extended by interacting the fixed effects with the subgroup indicator. For
each subgroup analysis, we tested whether the effect of DCS was heterogeneous by
evaluating the three-way interaction term of the DCS treatment x time x subgroup indi-
cator.
To compare DCS pre to DCS post, LMMs were repeated with three groups (DCS
pre, DCS post and placebo).
Results
Of the 57 included panic disorder with agoraphobia patients, 14% (n=8) dropped out
during therapy (see Figure 1). Mean ages differed between the study groups, with pa-
tients receiving DCS prior to an exposure session being significantly younger than the 4
other study groups (F (2,54):5.09,p:0.009), but there were no between group differences
on other socio-demographic variables at baseline, nor were there between-group differ-
ences on MI, BAI, BDI or PDSS scores.(Table 1 shows baseline characteristics). More-
over, comorbidity rates were similar between the three treatment conditions, as was the
use of antidepressants (including SSRIs, SNRIs and TCAs) and benzodiazepine.

Dropouts
There were no differences between completers and dropouts on the primary outcome
measure (the baseline MI ‘alone’ score), nor on socio-demographic or treatment condi-
tion parameters. However, the dropped out patients had significantly higher depression
scores at baseline than the completers (mean BDI scores in completers of 17.4 versus
26.6 in drop-outs (F(1.55):7.29,p:0.009).
DCS versus placebo

A within-time effect occurred with a 30% decrease in the MI “alone” score, with large
effect sizes: at six-months follow-up, δ= 0.9 (DCS) and δ= 0.8 (placebo).
Corroborating the analyses on the main outcome measure, a within-time effect was
reported on all secondary outcome measures (mean score on MI “accompanied”, BAI,
PDSS and BDI, Table 2), but no significant between-group differences in effect were
found between DCS and placebo (results not shown).
97
Chapter 4
Table 1
Baseline descriptives (n=57)
Socio-demographics Total DCS pre DCS post Placebo

(n=57) (n=19) (n=19) (n=19)
Age (years, SD) 35.4 (10.6) * 29.5 (6.2) 38.4 (11.3) 38.3 (11.4)
Gender (% female, n) 59.6 (34) 57.9 (11) 63.2 (12) 57.9 (11)
Education (years, SD) 13.3 (3.1) 12.7 (3.7) 13.1 (2.8) 14.2 (2.9)
Marital status (% married, n) 29.8 (17) 10.5 (2) 31.6 (6) 47.4 (9)
Primary outcome measure
MI ‘alone’ (mean score, SD) 2.9 (1.0) 2.8 (1.1) 3.2 (0.7) 2.8 (1.1)
Secondary outcome measures
MI ‘accompanied’(mean score, SD) 2.2 (0.8) 2.1 (0.9) 2.4 (0.7) 2.0 (0.7)
BAI (mean score, SD) 25.0 (10.9) 27.7 (10.2) 25.7 (11.9) 21.5 (10.2)
PDSS (mean score, SD) 14.9 (4.6) 15.4 (3.8) 14.6 (5.2) 14.7 (5.0)
BDI (mean score, SD) 18.7 (9.4) 18.1 (7.5) 19.7 (11.2) 18.4 (9.6)
Comorbidity
No comorbid disorder (%, n) 57.9 (33) 68.4 (13) 63.2 (12) 42.1 (8)
Current comorbid other anxiety disorder (%, n) 15.8 (9) 0 (0) 31.6 (6) 15.8 (3)
Current comorbid depressive disorder (%, n) 15.8 (9) 15.8 (3) 0 (0) 31.6 (6)
Medication
Use of antidepressant medication (% yes, n) 28.1 (16) 15.8 (3) 27.3 (5) 42.1 (8)
Use of benzodiazepines (% yes, n) 14.0 (8) 10.5 (2) 10.5 (2) 21.1 (4)
DCS pre: DCS augmentation half an hour prior to exposure sessions. DCS post: DCS augmentation
directly after exposure sessions.
MI: Mobility Inventory with a subscale ‘alone’ and a subscale ‘together with a significant other’. BAI:
Beck Anxiety Inventory, PDSS: the Panic Disorder Severity Scale, BDI: Beck Depression Inventory II.
* Significant difference in baseline age: F (2, 54): 5.09, p=0.009.
98
Table 2
Predicted means based on Linear Mixed Models for primary and secondary outcome measures at dif-
ferent times for DCS versus placebo (top part) and estimates for the condition*time interaction terms
and derived between group effect sizes of the primary outcome measure (bottom part).”
Predicted Condition T1 T2 T3 T4 T5 T6
mean score
MI ‘alone’ DCS (n=38) 3.03 2.93 2.56*** 2.30*** 2.18*** 2.18***
Placebo 2.77 2.41* 2.19*** 1.90*** 1.93*** 1.96***

(n=19)
MI DCS 2.27 2.15 1.87*** 1.69*** 1.63*** 1.63***

‘accompanied’
Placebo 2.03 1.83 1.58*** 1.40*** 1.41*** 1.43***
BAI DCS 26.68 26.24 20.89*** 17.75*** 19.31*** 18.73***
Placebo 20.53 19.81 17.95** 15.65*** 18.60*** 14.32***
PDSS DCS 14.97 10.28* 8.63*** 4.24*** 5.83*** 4.69***
Placebo 14.83 10.74*** 8.80*** 4.77*** 5.63*** 4.17***
BDI DCS 18.90 19.03 16.79 15.38* 14.62** 14.95**
Placebo 18.37 16.57 14.45 10.74*** 12.94** 12.24**
MI ‘alone’ Est ref 0.26 0.11 0.14 -0.01 -0.04
95% CI -0.07 - -0.24 - -0.21 - -0.36 - -0.39 - 4

0.60 0.46 0.49 0.34 0.32
t-test 1.56 0.61 0.81 -0.08 -0.20
p-value 0.12 0.54 0.42 0.94 0.85
ES -0.3 -0.1 -0.2 0.0 0.0
T1=Baseline;T2= Mid-study medication ;T3= post study-medication ;T4= post exposure therapy; T5=
follow up after 3 months; T6= follow up after 6 months.
Beck Anxiety Inventory, PDSS: the Panic Disorder Severity Scale., BDI: Beck Depression Inventory II.
Est: Estimate of the condition*time interaction term for the linear mixed model at different times. CI:
Confidential Interval. ES: effect size
significance level of within group effects: * P<0.05, **p<0.01, ***p<0.001
99
Chapter 4
Figure 2
Predicted means based on Linear Mixed Models for primary and secondary outcome measures.
directly after exposure sessions. T1=Baseline;T2= Mid-study medication;T3= post study-medication
;T4= post exposure therapy;T5= follow up after 3 months;T6= follow up after 6 months.
100
Table 3
Predicted means based on Linear Mixed Models for primary and secondary outcome measures at dif-
ferent times for DCS pre versus DCS post (top part) and estimates for the condition*time interaction
terms and derived between group effect sizes of the primary outcome measure (bottom part).”
Predicted mean Condition T1 T2 T3 T4 T5 T6

score
MI ‘alone’ DCS pre 2.81 2.67 2.48* 2.29* 2.23*** 2.17***

(n=19)
DCS post 3.23 3.19 2.63*** 2.31*** 2.31*** 2.22***

(n=19)
MI ‘t accompanied’ DCS pre 2.16 2.02 1.88* 1.77*** 1.71*** 1.73***
DCS post 2.36 2.29 1.87*** 1.61*** 1.56*** 1.56***
BAI DCS pre 27.26 27.26 22.57* 19.98** 23.19* 19.13**
DCS post 24.95 24.68 19.13* 15.66*** 14.45*** 15.37***
PDSS DCS pre 16.44 14.44 8.92*** 7.60*** 8.09*** 7.44***
DCS post 13.22 11.72 8.25*** 6.72*** 5.98*** 5.39***
BDI DCS pre 18.11 20.32 18.78 18.04 15.56 15.12
DCS post 19.68 17.74 14.98* 12.74** 13.70** 14.60*
MI ‘alone’
4
DCS pre vs DCS 95% CI ref -0.3- -0.68- -0.82- -0.95 - -0.80 -
post 0.49 0.13 0.01 -0.01 0.06
p-value 0.63 0.19 0.06 0.02 0.10
ES -0.2 -0.3 -0.4 -0.3 -0.2
DCS pre vs placebo 95% CI -0.09- -0.78- 0.02- -0.08- -0.18-

0.56 0.60 0.70 0.61 0.53
p-value 0.16 0.13 0.04 0.13 0.33
ES -0.2 -0.3 -0.4 -0.3 -0.2
DCS post vs placebo 95% CI -0.07- -0.42- -0.46- -0.67- -0.60-

0.70 0.38 0.36 0.14 0.20
p-value 0.11 0.92 0.81 0.20 0.33
ES -0.3 0.0 0.1 0.3 0.2
directly after exposure sessions.
T1=Baseline;T2= Mid-study medication ;T3= post study-medication ;T4= post exposure therapy; T5=
follow up after 3 months; T6= follow up after 6 months.
CI: Confidential Interval. ES: effect size
significance level of within group effects: * P<0.05, **p<0.01, ***p<0.001
101
Chapter 4
DCS augmentation pre versus post exposure sessions

A within-time effect was reported for both groups with large effect sizes: at six-months
follow-up, δ= 0.7 (DCS pre) and δ= 1.1 (DCS post). A between-group effect was
reported at three-months follow-up; the DCS post group showed greater symptom
reduction on the MI “alone” subscale, compared to the DCS pre group (95%CI -0.95
- -0.01,p:0.02) with an effect size of 0.3. However, no differences in effect were found
between DCS pre and placebo, nor between DCS post and placebo.
With regard to the secondary outcome measures, significant differences in effect
were found in favor of DCS augmentation post exposure relative to DCS augmentation
preceding exposure sessions. At three-months follow-up, the DCS post group showed
significantly larger reduction of anxiety as well as depressive scores as measured
with BAI resp. BDI-II compared to the DCS pre group (95%CI -12.6 - -0.18,p:0.04 for
BAI scores, 95%CI -12.7 - -1.1,p:0.02 for BDI II scores). However, no differences were
found on these measures between DCS administration pre, post and placebo (data not
shown).

Ancillary analyses
Additional analyses were conducted to investigate whether specific subgroups of pa-
tients experience a relative advantage of DCS administration. None of the mixed model
analyses of the subgroups with covariates: responders versus non-responders, early
versus late responders, and severely versus mildly affected patients yielded a time x
group interaction effect (data not shown).
Adverse effects
When patients reported any symptoms or complaints (measured with the Fawcett
side-effects list), the interviewer checked for a possible relationship to DCS. Four pa-
tients receiving DCS reported possible side-effects due to study medication (nausea,
fatigue), but none of these reports necessitated action. Thus, we conclude that no im-
portant adverse effects of DCS occurred during this study.
Discussion
To the best of our knowledge, this study has been the largest to investigate the added
value of DCS as an exposure treatment enhancer in patients with panic disorder with
agoraphobia and the first to investigate DCS augmentation pre exposure sessions ver-
sus DCS post exposure sessions.
We found no enhancement effect of 125 mg DCS augmentation of six exposure
sessions within a twelve-session exposure therapy protocol. Neither did ancillary anal-
yses in specific groups of patients according to symptom severity or speed of response
bring to light any preferential effects. The only two studies in panic disorder with/with-
out agoraphobia published to date showed more positive outcomes. Otto et al.22 per-
formed an RCT in which 31 patients with panic disorder with or without agoraphobia re-
ceived 50 mg of DCS or placebo one hour before three weekly interoceptive exposure
sessions. They reported that the DCS group showed a significantly greater reduction
of symptoms (effect size 1.1) directly post treatment In a second RCT, Siegmund et al.23
conducted a study of 39 patients with panic disorder with agoraphobia who underwent
102
eight sessions of group CBT, accompanied by three individual sessions of in-vivo expo-
sure (flooding). The latter three individual sessions were augmented with either 50mg
DCS or placebo. No additional effects of DCS were found, although post hoc analyses
yielded a trend for DCS to accelerate symptom reduction in patients at the severe end
of the spectrum.
However, we were unable to replicate any of these previous findings of an additional
effect of DCS in patients with panic disorder,22 nor could we replicate earlier subgroup
analyses indicating that DCS might speed up therapy or might be effective in patients
with more severe symptoms.23,24,35
The most pronounced differences between these previous studies and our study
are the dosage and duration of DCS augmentation: Otto el al.22 and Siegmund et al.23
used 50 mg of DCS prior to three exposure sessions whereas we used 125 mg of
DCS either pre or post six exposure sessions. Unfortunately, no dose-finding studies
have been performed in humans to date, but it appears that lower dosages might be
more likely to enhance treatment than higher dosages31 and that tolerance of DCS can
occur.52 Finally, differences in results between previous studies and our might be the
result of differences in study group characteristics. However, baseline measures of
core anxiety symptoms (on PDSS, BAI and MI) were highly comparable with this study,
baseline BDI scores in our sample were higher compared to the sample of Siegmund
et al.23 (18.7 versus 12.9). This might (partly) explain between-study differences.
This is the first study to compare DCS pre exposure with DCS post exposure in hu-
mans. One previous study27 compared DCS post versus placebo as an addition to two
exposure sessions in patients with acrophobia, but did not include DCS pre as a com- 4
parative treatment condition. In this study no additional effect of DCS was found. We
did find a favorable effect of DCS augmentation directly after exposure when compared
to administration preceding exposure, with an effect size of 0.3. However, the clinical
meaning of this result is yet to be determined, since neither DCS pre nor DCS post
showed enhancing effects when compared to placebo. However, our data modestly
suggest that DCS post exposure might be at least equally effective as DCS pre expo-
sure. This could be of clinical importance, since it offers the therapeutic advantage of
administrating DCS only after successful exposure and thereby limits the possible risks
of DCS; previous (animal) research has shown that DCS might also enhance sessions
in which anxiety conditioning instead of extinction has taken place.32

This study is the largest so far for panic disorder with agoraphobia, including 57 pa-
tients; it is the first to compare DCS prior to an exposure session versus DCS after an
exposure session. However, some limitations deserve mention: the sample is still rela-
tively small. Although daytime use of benzodiazepines served as an exclusion criterion
of the study, the use of benzodiazepines might have influenced exposure sessions, as
was shown in a previous study regarding PTSD.53 However, patients who continued
to use benzodiazepines did not differ in baseline characteristics and were randomly
distributed to the three treatment conditions. Finally, the use of 125 mg of DCS over
the six-week period might have led to some form of reduced effect and tolerance. How-
ever, two previous studies used dosages and durations comparable to our study and
103
Chapter 4
found enhancing effects from DCS.16,17 As long as studies on dosage and the length of
augmentation to reach an optimal effect of DCS have yet to be conducted, this issue
remains one of speculation.
In conclusion: although definite answers on the clinical relevance of DCS as a treat-
ment enhancer cannot be given, since questions with respect to optimal dosages and
the optimal number of sessions to enhance remain unanswered, we can conclude that
augmentation with DCS in patients with panic disorder and agoraphobia is not or -at
most- only very modestly effective, with some indication that that DCS, administered
after an exposure session might yield somewhat better results than when administered
preceding sessions.
104
Reference list
1. Kessler RC, Chiu WT, Demler O, Walters EE. Prevalence, Severity, and Comorbid-
ity of 12-month DSM-IV Disorders in the National Comorbidity Survey Replication.
Arch Gen Psychiatry 2005;62:617-709.
2. Furukawa TA, Watanabe N, Churchill R. Psychotherapy plus antidepressant for
panic disorder with or without agoraphobia: Systematic review. Br J Psychiatry
2006;188:305-312.
3. Watanabe N, Churchill R, Furukawa T. Combination of psychotherapy and ben-
zodiazepines versus either therapy alone for panic disorder: a systematic review.
BMC Psychiatry 2007;14:7-18.
4. Batelaan NM, van Balkom AJ, Stein DJ. Evidence-based pharmacotherapy of pan-
ic disorder: An update. Int J Neuropsychopharmacol 2012;15:403-415.
5. Bakker A, van Balkom AJ, Spinhoven P et al. Follow-up on the treatment of pan-
ic disorder with or without agoraphobia: a quantitative review. J Nerv Ment Dis
1998;186:414-419.
6. Cuijpers P, Sijbrandij M, Koole Sl et al. Adding psychotherapy to antidepressant
medication in depression and anxiety disorders: a meta-analysis. World Psychiatry
2014;13:56-67.
7. Brown TA, Barlow DH. Long-term outcome in cognitive-behavioral treatment of
panic disorder: Clinical predictors and alternative strategies for assessment. J
Consult Clin Psychol 1995;63:754−765.
8. Durham RC, Chambers JA, Power KG et al. Long-term outcome of cognitive be-
haviour therapy clinical trials in central Scotland. Health Technol Assess 2005;9:1- 4
174.
9. Pollack MH, Otto MW, Roy-Byrne PP et al. Novel treatment approaches for refrac-
tory anxiety disorders. Theoretical Review. Depress Anxiety 2008;25:467–476.
10. Amaral OB, Roesler R. Targeting the NMDA receptor for fear-related disorders.
Recent Pat CNS Drug Discov 2008;3:166-78.
12. Ledgerwood L, Richardson R, Cranney J. D-cycloserine facilitates extinction of
learned fear: effects on reacquisition and generalized extinction. Biol. Psychiatry
2005 57:841-847.
13. Walker DL, Ressler KJ, Lu KT, Davis M. Facilitation of conditioned fear extinction
by systemic administration or intra-amygdala infusions of D-cycloserine as as-
sessed with fear-potentiated startle in rats. J Neurosci 2002;22:2343-2351.
14. Bouton ME, Mineka S, Barlow DH. A modern learning theory perspective on the
etiology of panic disorder. Psychol Rev 2001;108:4-32.
15. Quirk GJ. Extinction: new excitement for an old phenomenon. Biol Psychiatry
2006;60:317-318.
16. Wilhelm S, Buhlmann U, Tolin DF et al. Augmentation of behavior therapy with D-cy-
closerine for obsessive-compulsive disorder. Am J Psychiatry 2008;165:335-341.
18. Storch EA, Merlo LJ, Bengtson M et al. D-cycloserine does not enhance expo-
105
Chapter 4
sure-response prevention therapy in obsessive-compulsive disorder. Int Clin Psy-

chopharmacol 2007;22:230-237.
19. Storch EA, Murphy TK, Goodman WK et al. A preliminary study of D-cycloserine
augmentation of cognitive-behavioral therapy in pediatric obsessive-compulsive
disorder. Biol Psychiatry 2010;68:1073-1076.
20. Guastella AJ, Richardson R, Lovibond PF et al. A randomized controlled trial of
D-Cycloserine Enhancement of Exposure therapy of social anxiety disorder. Biol
Psychiatry 2008;63:544-549.
21. Hofmann SG, Meuret AE, Smits JA et al. Augmentation of exposure therapy with
D-cycloserine for social anxiety disorder. Arch Gen Psychiatry 2006;63:298-304.
22. Otto MW, Tolin DF, Simon NM et al. Efficacy of D-cycloserine for enhanc-
ing response to cognitive-behavior therapy for panic disorder. Biol Psychiatry
2010;67:365-370.
chiatr Res 2011;45:1042-1047.
24. de Kleine RA, Hendriks G, Kusters WJC et al. A randomized placebo-controlled tri-
al of d-cycloserine to enhance exposure therapy for posttraumatic stress disorder.
25. Litz BT, Salters-Pedneault K, Steenkamp MM et al. A randomized placebo con-
trolled trial of D-cycloserine and exposure therapy for posttraumatic stress disor-
der. J Psychiatr Res 2012;46:1184-1190.
26. Ressler KJ. Cognitive enhancers as adjuncts to psychotherapy: use of D-cy-
closerine in phobic individuals to facilitate extinction of fear. Arch Gen Psychiatry
2004;61:1136-1144.
27. Tart CD, Handelsman PR, De Boer LB et al. Augmentation of exposure therapy
with post-session administration of D-cycloserine. J Psychiatr Res 2013;47:168-174.
28. Nave AM, Tolin DF, Stevens MC et al. Exposure Therapy, d-Cycloserine, and
Functional Magnetic Resonance Imaging in Patients With Snake Phobia: A Ran-
domized Pilot Study. J Clin Psychiatry 2012;73:1179-1186.
29. Norberg, MM, Krystal JH, Tolin DF. A Meta-Analysis of D-cycloserine and the facil-
itation of fear extinction and exposure therapy. Biol Psychiatry 2008;63:1118-1126.
30. Bontempo A, Panza KE, Bloch MH. D-cycloserine augmentation of behavioral
2012;73:533-537.
sure therapy for anxiety disorders in humans? A meta-analysis. PLoS ONE
2014;9:e93519.
231.
34. Smits JA, Rosenfield D, Otto MW et al. D-cycloserine enhancement of exposure
106
J Psychiatr Res 2013;47:1455-1461.
35. Hofmann SG, Smits JA, Rosenfield D et al. D-Cycloserine as an augmentation
strategy with cognitive-behavioral therapy for social anxiety disorder. Am J Psychi-
atry 2013;170:751-758.
36. First MB, Spitzer RL, Gibbon M, Williams JB. Structured clinical interview for DSM-
IV-TR Axis I Disorders, Patient Edition (SCID-I/P, version 2.0). New York: Biomet-
rics Research, New York State Psychiatric Institute;1996.
37. Groenestijn MA, van Akkerhuis GW, Kupka RW et al. Gestructureerd klinisch inter-
view voor de vaststelling van DSM-IV As I stoornissen. Lisse:Swets test Publish-
ers;1997.
38. Beck AT, Steer RA, Brown GK. Manual for Beck Depression Inventory-II. San An-
tonio, TX: Psychological Corporation;1996.
39. Schmand B, Lindeboom J, van Harskamp F. NLV | Nederlandse Leestest voor Vol-
wassenen, Handleiding. Lisse: Swets & Zeitlinger ;1992.
40. Chambless DL, Caputo GC, Jasin SE et al. The Mobility Inventory for Agorapho-
bia. Behav Res Ther 1985;23:35-44.
41. Chambless DL, Sharpless BA, Rodriquez D et al. Psychometric Properties of the
Mobility Inventory for Agoraphobia: Convergent, Discriminant, and Criterion Relat-
ed Validity. Behav Ther 2011;42:689-699.
42. Beck AT, Epstein N, Brown G, Steer RA. An inventory for measuring clinical anxi-
ety: psychometric properties. J Consult and Clin Psychol 1988;56:893-897.
43. Shear MK, Brown TA, Barlow DH et al. Multicenter collaborative panic disorder se- 4
verity scale. Am J Psychiatry 1997;154:1571–1575.
44. Shear MK, Rucci P, Williams J et al. Reliability and validity of the Panic Disorder
Severity Scale: Replication and extension. J Psychiatr Res 2001;35:293-296.
46. Beck AT, Steer RA, Ball R, Ranieri W. Comparison of Beck Depression Inventories
-IA and -II in psychiatric outpatients. J Pers Assess 1996;67:588-5597.
47. Richter P, Werner J, Heerlein A et al. On the Validity of the Beck Depression Inven-
tory. A review. Psychopathology 1998;31:160-168.
48. Fawcett J. Fawcett Side Effect Scale. Psychopharmacol Bull 1987;3:322-324.
49. Clark D.M. A cognitive approach to panic. Behav Res Ther 1986;24:461–470.
50. Craske MG, Barlow DH. Mastery of your anxiety and panic: Therapist guide, 4th
ed. New York: Oxford University Press;2007.
51. Morris SB. Estimating Effect Sizes From the Pretest-Posttest-Control Group De-
signs. Organizational Research Methods 2008;11:364-386.
52. Quartermain D, Mower J, Rafferty MF et al. Acute but not chronic activation of the
NMDA-coupled glycine receptor with D-cycloserine facilitates learning and reten-
tion. Eur J Pharmacol 1994;257:7-12.
53. Rothbaum BO, Price M, Jovanovic T et al. A Randomized, Double-Blind Evaluation
of D-Cycloserine or Alprazolam Combined With Virtual Reality Exposure Therapy
for Posttraumatic Stress Disorder in Iraq and Afghanistan War Veterans. Am J
Psychiatry 2014;171:640–648.
107
108
Chapter 5
SUMMARY AND GENERAL DISCUSSION
109
Chapter 5
SUMMARY AND GENERAL DISCUSSION

This thesis aims to address lacunas in the current knowledge of complex anxiety dis-
orders. This is an imperative topic since complex anxiety disorders tend to develop a
chronic course1 and because current guidelines are incomplete. Whereas the guideline
for depression emphasise the impact and treatment options for chronicity or treatment
resistance, there is a lack of information on this topic in the anxiety guidelines (www.
ggzrichtlijnen.nl).
In this thesis, several studies are presented regarding the various aspects of com-
plex anxiety disorders. This includes research on a putative risk factor for complex anx-
iety disorders (comorbidity), on a potential underlying mechanism of complex anxiety
disorders (the tendency to overestimate threat; expectancy bias), and on a potential
new treatment enhancer (d-cycloserine; DCS).
In this chapter, the main findings of our studies are summarised. Furthermore, the
methodological limitations of our findings, the clinical implications, and recommenda-
tions for future research are discussed. Finally, we offer concluding statements, which
place our findings in a broader prospective.
Risk factor for complex anxiety disorders; comorbidity

Introduction
As emphasised in the introduction, there is some ambiguity regarding the role of co-
morbidity in clinical decision making. On the one hand, comorbidity in anxiety disorders
is common2-5 and recognised as independent risk factor for chronicity.1 On the other
hand, the relevance of comorbidity in clinical practice is kept underexposed. This lack
of knowledge might be part of the reason why clinicians appear reluctant to address
all comorbid disorders; they may doubt the relevance of comorbidity in the treatment
of the primary anxiety disorder. This reluctance is reinforced by the multidisciplinary
guidelines where chronicity and comorbidity in anxiety disorders receive little attention
(www.ggzrichtlijnen.nl). In addition, this reluctance is fuelled by the current debate
about the origin of psychiatric disorders. For example, comorbidity in psychiatry is
based on the DSM classifications (with overlapping symptom clusters) rather than
based on the co-existence of two distinct diseases (as the original definition of comor-
bidity requires).6 It has been suggested that comorbidity might reflect different manifes-
tations of a single ‘core’ disease of which the etiology is currently unknown.7 However,
as long as psychiatric disorders are not based on neurobiological or genetic etiology,
it is recommended to use the DSM classification to diagnose psychiatric disorders and
comorbidity.8 This thesis aim to address clinical implications of DSM IV diagnosed co-
morbidity regardless of the etiology of this type of comorbidity.
In Chapters 2.1, 2.2, and 2.3, we present three studies regarding comorbidity in
anxiety disorders. These three studies were based on samples derived from two longi-
tudinal studies, which are the Netherlands Study of Depression and Anxiety (NESDA)
and the Netherlands OCD Association (NOCDA) study. The NESDA study encompass-
es a naturalistic cohort study of 2,981 participants, recruited from general population,
general practices, and mental health organisations in the Netherlands.9 The NOCDA
cohort entails an ongoing naturalistic longitudinal cohort study conducted among 419
adult patients with a lifetime OCD.10
110
Summary and general discussion
Chapter 2.1
In chapter 2.1, we present the first study in which we investigated the clinical relevance
of diagnosing type of comorbidity and the number of comorbid disorders in anxiety dis-
orders.
Using a sample of 1004 NESDA participants with at least one 12-month anxiety
disorder, the following four groups were compared to investigate comorbidity in anxi-
ety disorders: single anxiety disorder, anxiety-anxiety comorbidity, anxiety-depressive
comorbidity, and “double” comorbidity (i.e. anxiety and depressive comorbidity). To di-
agnose DSM-IV anxiety disorders and depressive disorders in the past 12 months, the
Dutch version of the Composite International Diagnostic Interview (CIDI, version 2.1)
was used.11 We investigated sociodemographics, vulnerability factors, and clinical char-
acteristics to determine whether the type and number of comorbid disorders in anxiety
disorders were relevant.
Results clearly demonstrate that comorbidity in anxiety disorders is associated with
higher severity and more chronicity compared with single anxiety disorders. Moreover,
the results suggest the importance to distinguish the type and the number of comorbid
disorders, since different patterns are observed. For instance, anxiety-anxiety comor-
bidity is associated with an earlier age of onset and a more chronic course, whereas
anxiety-depressive comorbidity has a longer duration of treatment and leads to im-
paired functioning. Furthermore, “double” comorbidity leads to even higher severity,
chronicity, and impairment.
These findings are in line with earlier research on anxiety-anxiety comorbidity and
severity of symptoms,12 and large studies regarding anxiety-depressive comorbidity.13-17
We conclude with the recommendation that in clinical practice, all comorbid disor-
ders present in patients with an anxiety disorder should be diagnosed. Given the higher
severity, unfavourable course, and greater impairments, patients with an anxiety disor- 5
der and comorbidity should receive prompt and adequate treatment.
Chapter 2.2
In chapter 2.2, the second study is presented in which we explored many clinical fac-
tors and risk indicators associated with comorbidity in OCD. We used a large sample
of 382 patients with a current DSM-IV diagnosis of OCD from the NOCDA sample, as-
sessed with the Structured Clinical Interview for DSM-IV-TR (SCID-I/P).18
We found that comorbidity, especially of multiple comorbid disorders, was associated
with more chronicity, more severe OCD, and more negative consequences on daily life.
These results add to current knowledge by providing a possible explanation for ear-
lier inconclusive findings; whereas some studies did not find any association between
comorbidity and OCD severity3 or chronicity,19 other studies did discover this connec-
tion.20,21 We initially found that the severity of OCD did not differ substantially between
‘pure’ OCD and OCD with comorbid disorders. Nevertheless, when dividing the group of
comorbid disorders into OCD with one comorbid disorder and OCD with two or more co-
morbid disorders, it appeared that the latter group of patients suffered from significantly
more severe and chronic OCD. The same pattern was found for an association between
comorbidity and OCD chronicity; when differentiating between OCD and one or multiple
comorbid disorders, it appeared that multiple comorbidity was associated with more
111
Chapter 5
chronicity. These results may explain the inconsistencies in the current literature.
We conclude with the recommendation that comorbidity should be taken into ac-
count when planning treatment; for example, a comorbid severe depression may be an
indication of antidepressants.
Chapter 2.3
In chapter 2.3, the third study regarding comorbidity is presented. We investigated the
clinical relevance of comorbid OC-symptoms in a large sample of NESDA participants
(n=2125) with current or remitted anxiety and depressive disorders and healthy con-
trols. As mentioned in the introduction, DSM-IV diagnosis of OCD was not included in
the baseline interview of NESDA. In the two year follow-up wave, OC-symptoms were
first assessed using The Young Adult Self-Report Scale for obsessive compulsive
symptoms (YASR-OCS),22 which is a self-report instrument suitable as a screener for
the presence of OC-symptoms, yet not to establish a definite DSM-IV diagnosis of OCD.
The results indicate that clinically relevant OC-symptoms are present in a substan-
tial proportion of patients with anxiety and depressive disorders and are associated with
higher severity. Additionally, clinically relevant OC-symptoms predicted first onset and
relapse of anxiety and depressive disorders. Moreover, clinically relevant OC-symp-
toms predicted chronicity in patients with an anxiety and depressive disorder, but not in
patients with an anxiety or depressive disorder. We conclude with the recommendation
that clinically relevant OC-symptoms might warrant treatment and should alert clinicians
for unfavourable outcomes and relapse in anxiety and depressive disorders.
Methodological considerations
For the studies presented in Chapter 2, data were drawn from the Netherlands Study of
Depression and Anxiety and the Netherlands OCD Association study. Two large-scaled
longitudinal cohort studies aimed at examining the long-term course of anxiety, de-
pression, and OCD. The main limitations of the studies presented in Chapters 2.1 and
2.3 was the lack of a DSM-IV diagnosis of OCD. This was not included in the baseline
interview of NESDA and, therefore, OCD could not be included in Chapter 2.1, and in
Chapter 2.3 we needed to use the YASR-OCS as a screening instrument for OC-symp-
toms. Participants with scores above the cut-off point of 7 were regarded as suffering
from clinically relevant OC-symptoms (and not definite OCD). This cut-off point was
based on earlier studies.23
Another limitation of chapters 2.1 and 2.2 was the cross-sectional design. A pro-
spective design would be more appropriate to assess the outcome and impact of co-
morbidity on the course of anxiety disorders. Course and, thus, chronicity were now
established retrospectively and specific pathways by which comorbidity may be related
to course and chronicity could not be clarified.
Overall conclusions regarding current clinical practice

With the current debate and ongoing research on possible shared etiology between
anxiety and depressive disorders, the concept of comorbidity will remain subject to
debate. However, our studies demonstrate the clinical relevance of comorbidity regard-
less of the etiology of comorbidity. This finding leads to a number of recommendations
112
for clinical practice, as mentioned in the previous paragraphs. In short, we recommend

awareness of comorbidity and adequate treatment. In this paragraph, we put these rec-
ommendations in a broader perspective.
First, awareness of comorbidity is recommended; for example, in the case of OC
symptoms. OC symptoms (but not a definite diagnosis of OCD) might not always war-
rant treatment themselves, yet are still a valuable specifier for anxiety and depressive
disorders and direct treatment. These findings are comparable with the presence of
anxiety symptoms in depression or with the presence of panic attacks across the full
range of psychopathology.24,25 This knowledge could add to models for profiling and
staging of anxiety and depressive disorders.26-28 We recommend adding comorbidity of
OC symptoms to clinical staging models in line with other predictors of chronicity such
as severity, level of functioning, and illness duration.29,30 Furthermore, the algorithms
that are currently being developed may improve clinical prediction and direct treat-
ment.
Second, we suggest that awareness of comorbidity early in the diagnostic and
treatment process will improve and guide treatment. In the current treatment guide-
lines, comorbidity is not taken into account in the first three steps of the decision-mak-
ing process. In other words, the first three treatment options are currently advised
regardless of comorbid disorders. For this reason, comorbidity should be considered
earlier in the treatment process.
Third, we recommend adequate treatment. But what is adequate treatment of co-
morbidity in anxiety disorders? Does this include pharmacological, psychological or
combined interventions? Current guidelines suggest either pharmacological or psycho-
logical treatments for anxiety disorders and combined treatment is only recommended
when anxiety disorders are complicated by a major depression (www.ggzrichtlijnen.
nl). This statement is still subject to debate. No recent literature has added to current 5
knowledge on how to treat comorbidity. In contrast, comorbidity is often an exclusion
criterion for participation in a randomised controlled trial and most recent meta-analy-
ses also rule comorbidity out.31 Thus, no evidence-based statements for the pharmaco-
logical treatment of comorbidity can be made at this time. There is an interesting devel-
opment regarding a new form of combined treatment, which is psychological treatment
enhancement by new drugs. One of these enhancers (d-cycloserine) is discussed in
the final chapter of this thesis.
There are some issues regarding psychological treatment that deserve mentioning.
First, the influence of comorbid disorders on treatment outcome is a subject of debate.
Although some studies have associated comorbidity with worse treatment outcome,32-34
there is growing support that - taken into account that comorbidity might be associated
with higher pre-treatment severity of the primary anxiety disorder - comorbidity is not
associated with worse treatment outcome.35,36 Thus, so far, the question whether the
presence of comorbidity might direct to a different psychological treatment remains
unanswered.
Second, some research suggests that by treating the primary diagnosis, comorbid
symptoms also decline.36,37 Although this might be the case, the presence of residual
symptoms (not a full recovery of the comorbid disorder) is an important risk factor for
relapse of the comorbid disorder.32
113
Chapter 5
Finally, it should be noted that simultaneous application of more than one disor-
der-speciﬁc CBT does not seem to enhance treatment outcome.38
Considering these issues, current research focusses on a new form of CBT caled
‘transdiagnostic treatment’. This form of CBT targets more than one diagnosis by, for
example, focusing on regulating emotions.39 Some studies have demonstrated support
for the transdiagnostic treatment of anxiety disorders.40-42 Direct comparison of a trans-
diagnostic treatment with a disorder-specific CBT showed no advantages of the trans-
diagnostic treatment on primary and comorbid disorders.43 In sum, we can conclude
that there is preliminary evidence for the efficacy of transdiagnostic CBT that aims
to treat symptomatic comorbidity,44 but more research is required to improve current
transdiagnostic protocols and to compare disorder-specific treatment with transdiag-
nostic treatment.45
Recommendations for future research

Research on pharmacological treatment of comorbidity is needed and large ran-
domised controlled trials should not exclude patients with comorbidity, yet explore
pharmacological treatment options for patients with comorbidity. Furthermore, trans-
diagnostic treatment is an interesting and promising psychological treatment strategy
for comorbidity. However, results are only preliminary and direct comparing of one
disorder-specific treatment versus non-specific group treatment is needed to evaluate
the effect of this transdiagnostic treatment on both the primary disorder as comorbid
disorders. When proven effective, non-specific group treatment could lead to several
potential improvements such as more variety in treatment options, more cost effective-
ness treatments, and better treatment of comorbid disorders.
Underlying mechanism of complex anxiety disorders; expectancy bias

Introduction
The study presented in Chapter 3 aimed to enhance current knowledge on a potentially
underlying mechanism of complex anxiety, which is the expectancy bias (the tendency
to overestimate the association between fear-(ir)relevant stimuli and the aversive stim-
uli). This study aimed to i) replicate earlier findings that anxiety patients tend to have
an expectancy bias; and ii) investigate the predictive value of expectancy bias towards
exposure therapy outcome in patients with panic disorder and agoraphobia.
This study was part of the RCT presented and discussed in Chapter 4, examining
the added value of DCS augmentation in patient with panic disorder with agorapho-
bia. For this expectancy bias study, we included 71 patients with panic disorder and
agoraphobia (receiving exposure therapy with or without DCS augmentation) and 65
matched healthy control subjects.
During the expectancy bias task, participants were exposed to a series of panic-re-
lated and neutral words and rated for each word on their expectancy that it would be
followed by a loud sound (the unconditioned stimulus). Results demonstrate i) higher
expectancy ratings to panic-related words compared with neutral words in both patients
and controls; and ii) higher expectancy ratings to both panic-related and neutral words
in patients compared with controls. These findings are in line with previous findings46-49
and suggest that patients with panic disorder and agoraphobia tend to overestimate
114
the association between fear-relevant stimuli (panic-related words) and fear-irrelevant

stimuli (neutral words). In other words, patients tend to overestimate threat on both anx-
iety- and neutral cues. This expectancy bias seems to be crucial in the development
of an anxiety disorder; (normal) fear may induce an expectancy bias, which, in turn,
enhances perceived threat. Perceived threat induces more fear and eventually leads to
an anxiety disorder.50
This is the first study to investigate the predictive value of expectancy bias on
treatment outcome in patients with a panic disorder and agoraphobia. Based on one
earlier study on spider phobia,50 we hypothesised that higher expectancy ratings before
treatment would be associated with worse treatment outcome. However, our results
demonstrates that higher expectancy rates do not predict treatment outcome in the
patient group.
The main limitation of this study, as said before, was the use of subjective US expec-
tancy ratings only. This could possibly explain the difference in results between our
study and the spider phobia study as more objective psychophysiological outcome
measures might be more sensitive.
Another limitation of the study was that no post-treatment expectancy bias task was
assessed. Adding a post-treatment task might shed further light on the possibility that
expectancy bias is down-regulated as a result of treatment.
Overall conclusions regarding current clinical practise

This part of the thesis has more experimental research. Nonetheless, some clinical
implications can be mentioned. First, therapists should be aware of the presence of
expectancy bias in patients with panic disorder and agoraphobia. The tendency to 5
overestimate threat should be mentioned and addressed in psychotherapy. Second,
it deserve mentioning that understanding the different mechanisms of fear acquisition
and extinction in patients with an anxiety disorder might potentially lead to better treat-
ment options. Exposure-based therapy is an example of experimental research as the
basis for treatment of anxiety disorders and continues to change due to new insights
into underlying mechanisms. The tendency to overestimate threat is a possible expla-
nation of (part of) the development of an anxiety disorder, a chronic course, and the
development of treatment resistance.

Future research should focus on a more sensitive expectancy bias task with objective
and subjective outcome to investigate the possible association between expectancy
bias and treatment outcome, adding a post-treatment measurement as well as psycho-
physiological outcome parameters. Such a sensitive expectancy bias task might serve
as a reliable predictor for treatment outcome, and, theoretically, eventually might serve
as a reliable task that can predict if - and what type of - treatment will be effective.
Treatment enhancement in anxiety disorders; d-cycloserine

Although current treatment for anxiety disorders is effective for at least a part of the
115
Chapter 5
patients with an anxiety disorder, there is still room for improvement. This is especially
true from a translational point of view; pharmacological medicine based on neurobio-
logical underpinnings of anxiety disorders. In Chapter 4, we present the RCT we per-
formed to investigate the additional effect of d-cycloserine (DCS) in exposure therapy
for patients with a panic disorder with agoraphobia.
As mentioned in the introduction of this thesis, all studies investigating DCS have
used different procedures, different dosages, and different timing and frequency of
augmentation, with small sample sizes and different types of anxiety disorders. We de-
signed a randomised, double-blinded, placebo-controlled trial to investigate the addi-
tional effect of DCS to exposure therapy. For this study, we chose patients with a panic
disorder with agoraphobia. Our hypothesis was that this ‘classic’ anxiety disorder, with
severe patients and a treatment protocol with exposure elements in every sessions
augmented with DCS, would be a good setting for our research question.
Our second research question was whether DCS prior to exposure sessions would
be more effective compared with DCS after exposure. This study is the first human
study to compare DCS prior directly with DCS after exposure therapy, and animal re-
search has shown that DCS after exposure is equally effective compared with DCS
prior to exposure.51,52 The timing of DCS is clinically important since animal studies have
shown that DCS might also enhance unsuccessful exposure sessions,53 and DCS might
only have added value when combined with a successful session.54,54 Therefore, inves-
tigating the efficacy of DCS augmentation after exposure sessions is highly relevant.
We tested 57 patients from three Dutch institutions who received 125 mg of DCS
or placebo as addition to 6 exposure sessions within a 12-session exposure proto-
col. Moreover, 19 patients received DCS prior to exposure, 19 patients received DCS
after exposure, and 19 patients received placebo. We found no differences between
DCS and placebo, nor could we replicate earlier subgroup analyses indicating that
DCS might speed up therapy or might be effective in patients with more severe symp-
toms.56-58 Neither DCS pre- or DCS post exposure showed additional improvement
of exposure therapy compared with placebo. We found small significant differences
however, in favour of DCS augmentation after exposure compared with DCS prior to
exposure (effect size 0.3). The clinical meaning of this finding is yet to be determined,
since neither DCS pre nor DCS post showed enhancing effects compared with pla-
cebo. Nonetheless, the effect size we found is comparable to other studies and three
meta-analyses showing significant but modest additional effects of DCS in exposure
therapy in different types of anxiety disorders with effect sizes between 0.3 and 0.4.59-
61
The fact that our data could not replicate these findings could be explained by the
following methodological limitations: small sample size, suboptimal dosage or timing of
DCS and use of concomitant medication. These possible limitations are discussed in
the succeeding paragraph.
As previously stated, some possible limitations of our study design need to be ad-
dressed. The first limitation is sample size. Although we included the largest sample so
far with panic disorder and agoraphobia, we used three groups to compare DCS prior
and after exposure with placebo, which lowers statistical power.
116
Another limitation is the dosage of DCS augmentation. A recent meta-analysis has

suggested that low dosage of DCS (50 mg) might be more effective than higher dos-
ages (up to 500 mg).61 However, a previous meta-analysis did not find this preferential
effect of lower dosages.60 We used 125 mg of DCS (a moderately low dosage) possibly
leading to tolerance and a suboptimal effect. Conversely, two previous studies in which
dosages and number of sessions enhanced were comparable to our study did find en-
hancing effects of DCS.62,63 These limitations cannot be completely ruled out as partial
causes of our ‘null’ finding; yet, as long as dosage and timing have not been systemati-
cally evaluated, these issues remain speculative.
The final limitation is concomitant medication. A study including 128 patients with
OCD found that DCS might interact with antidepressant drugs leading to diminished
effects of DCS in patients using antidepressants.64 This interaction between DCS and
concomitant medication was not found in the meta-analysis by Rodrigues.61 In addition,
the use of benzodiazepines might influence the efficacy of exposure sessions.65 To rule
out these possible interaction effects, we performed post-hoc analyses on a sub-sam-
ple of antidepressant and benzodiazepine naive patients (n=37). The same results
were found; thus, no additional effects of DCS on treatment outcome are found.
Conclusions regarding clinical practise

Evidence for the use of DCS in clinical practice yields inconclusive results according
to studies conducted so far. However, data from meta-analyses are promising, and,
as our data modestly suggest, DCS post exposure sessions might be at least equally
effective as DCS pre exposure sessions. This could be of clinical importance, since it
offers the therapeutic advantage of administrating DCS only after successful exposure
sessions and, subsequently, limiting the possible risks of DCS to enhance unsuccess-
ful exposure sessions. 5
DCS is the most investigated and most promising cognitive enhancer up to date.
Nonetheless, there are more examples of translational research in which new pharma-
cological enhancement options have been developed based on the current knowledge
of anxiety disorders. Yohimbine, a noradrenergic enhancer (α2-adrenoceptor antago-
nist),66 brain-derived neurotrophic factor (BDNF),67 and cannabidiol68 are also examples
of possible treatment enhancers.

In spite of unanswered questions regarding optimal timing, dosages, and frequency of
augmentation, DCS seems a promising and interesting topic for more future research.
Ongoing research on pharmacokinetics and dynamics of DCS, as well as on the exact
working mechanism of DCS (including its action on the NMDA receptor), is required to
improve DCS application and to expand knowledge on the development and mainte-
nance of anxiety disorders. Finally, a longer follow-up, a large patient sample, and cost
effectiveness studies are needed to establish if DCS deserves a place in future guide-
lines and clinical practice.
Concluding remarks
This thesis includes research on the following major aspects of complex anxiety disor-
117
Chapter 5
ders: comorbidity as an important risk factor, expectancy bias as a potential underlying

mechanism, and d-cycloserine as a promising new treatment enhancement strategy.
In addition, the thesis includes the important issues regarding these three aspects.
We conclude that some major improvement in understanding risk factors for complex
anxiety disorders, underlying mechanism, and potential treatment enhancers have
been made during the past decade. All these aspects add to a better understanding
and better treatment of complex anxiety disorders. In sum, comorbidity should be taken
seriously and should guide treatment choices, the tendency to overestimate the associ-
ation between fear-(ir)relevant stimuli and aversive stimuli might be a reliable predictor
of treatment response, and DCS is a promising potential augmentation strategy for suc-
cessful exposure sessions.
118
Reference list
1. Penninx BW, Nolen WA, Lamers F et al. Two-year course of depressive and anx-
iety disorders: Results from the Netherlands Study of Depression and Anxiety
(NESDA). Journal of Affective Disorders 2011;33:76–85.
2001;110:585-599.
ple of patients with obsessive-compulsive disorder. J Affect Disord 2004;80:155-
162.
4. Lamers F, Van Oppen P, Comijs HC et al. Comorbidity patterns of anxiety and de-
try 2006;67:703-711.
6. Fernstein AR. The pre-therapeutic classification of co-morbidity in chronic diseas-
es. J Chronic Dis 1970;23:455–468.
7. Maj M. The aftermath of the concept of ‘psychiatric comorbidity’.Psychother Psy-
chosom 2005;74:67-68.
8. Van Oudheusden, LJ, Meynen, G, van Balkom AJ. Psychiatric comorbidity: theo-
retical and clnical dilemmas. Submitted at Tijdschrift voor psychiatrie 2015.
9. Schuurmans J, van Balkom AJ, van Megen HJ et al. The Netherlands OCD Asso-
chiatr Res 2012;21:273-85.
10. Penninx BW, Beekman AT, Smit JH et al. The Netherlands Study of Depression 5
Res 2008;17:121-140.
11. World Health Organization. Composite International Diagnostic Interview (CIDI):
version 2.1. Geneva, Switzerland; 1997.
12. Cougle JR, Timpano KR, Sachs-Ericsson N et al. Examining the unique relation-
ship between anxiety disorders and childhood physical and sexual abuse in the
National Comorbidity Survey-Replication. Psychiatry Res 2010;177:150-155.
1991;84:446-452.
chol Med 1997;27:861-873.
Arch Gen Psych 2003;60:993-1000.
119
Chapter 5
2005;87:43-55.
19. Eisen JL, Goodman WK, Keller MB et al. Patterns of remission and relapse in
obsessive-compulsive disorder: a 2-year prospective study. J Clin Psychiatry
1999;60:346-351.
20. Jakubovski E, Diniz JB, Valerio C et al. Clinical predictors of long-term outcome in
obsessive-compulsive disorder. Depression and Anxiety 2012;00:1-10.
21. Tükel R, Polat A, Özdemir Ö et al. Comorbid Conditions in Obsessive-Compulsive
Disorder. Compr Psychiatry 2002;43:204-209.
22. Achenbach TM. Manual for the Young Adult Self Report and Young Adult Behavior
Checklist. Department of Psychiatry, University of Vermont. Burlington VT; 1997.
23. Grootheest van DS, Cath DC, Beekman AT, Boomsma DI. Genetic and envi-
ronmental influences on obsessive-compulsive symptoms in adults: a popula-
tion-based twin-family study. Psychol Med. 2007;37:1635-1644.
orders, 4th edn. Washington DC: American Psychiatric Association;1994.
25. Batelaan NM, Rhebergen D, de Graaf R et al. Panic Attacks as s Dimension of
Disorders and Level of Functioning. J Clin Psychiatry 2012;73:1195-1202.
26. Balkom van AJ, Oosterbaan DB, Batelaan N et al. The characterization of anxiety
atrie 2012;11:935-940.
27. Fava GA, Tossani E. Prodromal stage of major depression. Early Intervention in
Psychiatry 2007;1: 9-18.
Sir Martin Roth. Psychother Psychosom 2008;77:133-138.
29. Batelaan NM, Rhebergen D, Spinhoven P et al. Two-Year Course Trajectories of
Anxiety Disorders: Do DSM Classifications Matter? J Clin Psychiatry 2014;75:985–
993.
30. Cosci F, Fava GA. Staging of mental disorders: systematic review. Psychother Psy-
chosom 2013;82:20–34.
31. Bandelow B, Reitt M, Röver C et al. Efficacy of treatments for anxiety disorders: a
meta-analysis. Int Clin Psychopharmacol 2015;30:183-192.
32. Brown TA, Antony MM, Barlow DH Diagnostic comorbidity in panic disorder: effect
on treatment outcome and course of comorbid diagnoses following treatment. J
Consult Clin Psychol 1995;63:408-418.
33. Steketee G, Eisen J, Dyck I et al. Predictors of course in Obsessive Compulsive
disorder. Psychiatry Res 1999;89:229-238.
34. Tsao JC, Lewin MR, Craske MG. The effects of cognitive-behavior therapy for pan-
ic disorder on comorbid conditions. J Anxiety Disord 1998;12:357-371
35. Tsao JC, Mystkowski JL, Zucker BG, Craske MG. Effects of cognitive-behavioral
therapy for panic disorder on comorbid conditions: Replication and extension. Be-
hav Res Ther 2002;33:493–509.
120
36. Allen LB, White KS, David H. Barlow DH et al. Cognitive-Behavior Therapy (CBT)
for Panic Disorder: Relationship of Anxiety and Depression Comorbidity with Treat-
ment Outcome. J Psychopathol Behav Assess 2010;32:185–192.
37. Noyes R. Comorbidity in generalized anxiety disorder. Psychiatr Clin North Am
2001;24:41-55.
38. Craske MG, Farchione TJ, Allen LB et al. Cognitive behavioral therapy for panic
disorder and comorbidity: more of the same or less of more? Behav Res Ther
2007;45:1095-1109.
39. Barlow DH, Allen LB, Choate ML. Toward a Unified Treatment for Emotional Disor-
ders. Behavior Therapy 2004;35:205—230.
40. Erickson DH. Group cognitive behavioural therapy for heterogeneous anxiety dis-
orders. Cogn Behav Ther 2003;32:179-186.
41. Erickson DH, Janeck AS, Tallman K. A cognitive-behavioral group for patients with
various anxiety disorders. Psychiatr Serv 2007;58:1205-1211.
42. Norton PJ. An open trial of a transdiagnostic cognitive-behavioral group therapy for
anxiety disorder. Behav Ther 2008;39:242-250.
43. Craske MG, Niles AN, Burklund LJ et al. Randomized controlled trial of cognitive
behavioral therapy and acceptance and commitment therapy for social phobia:
outcomes and moderators. J Consult Clin Psychol 2014;82:1034-1048.
44. Rector NA, Man V, Lerman B. The Expanding Cognitive-Behavioural Therapy
Treatment Umbrella for the Anxiety Disorders: Disorder-Specific and Transdiag-
nostic Approaches. Can J Psychiatry 2014;59:301–309.
sion. Depression and Anxiety 2012;29:749-753.
46. Amrhein C, Pauli P, Dengler W, Wiedemann G. Covariation bias and its physiologi-
cal correlates in panic disorder patients. Journal of Anxiety Disorders 2005;19:177- 5
191.
47. Pauli P, Montoya P, Martz G. Covariation bias in panic-prone individuals. Journal of
Abnormal Psychology 1996;105:658-662.
48. Pauli P, Montoya P, Martz G. On-line and a posteriori covariation estimates in pan-
ic-prone individuals: Effects of a high contingency of shocks following fear irrele-
vant stimuli. Cognitive Therapy and Research 2001;25:23-36.
disorder. Journal of Anxiety Disorders 2001;15:401-412.
50. de Jong PJ, van den Hout MA, Merckelbach H. Covariation bias and the return of
fear. Behaviour Research and Therapy 1995;33:211-213.
231.
54. Smits JA, Rosenfield D, Otto MW, et al. D-cycloserine enhancement of exposure
121
Chapter 5
J Psychiatr Res 2013;47:1455-1461.

55. Smits Ja, Rosenfield D., Otto MW, et al. D-cycloserine enhancement of fear Ex-
tinction is specific to succesfull exposure sessions: Evidence from the treatment of
height phobia. Biol psychiatry 2013;73:1054-1058.
chiatr Res 2011;45:1042-1047.
57. de Kleine RA, Hendriks G, Kusters WJ et al. A randomized placebo-controlled trial
of d-cycloserine to enhance exposure therapy for posttraumatic stress disorder.
58. Hofmann SG, Smits JA, Rosenfield D et al. D-Cycloserine as an augmentation
strategy with cognitive-behavioral therapy for social anxiety disorder. Am J Psychi-
atry 2013;170:751-758.
59. Norberg, MM, Krystal JH, Tolin DF et al. A Meta-Analysis of D-cycloserine and the
facilitation of fear extinction and exposure therapy. Biol Psychiatry. 2008;63:1118-
1126.
60. Bontempo A, Panza KE, Bloch MH et al. D-cycloserine augmentation of behavioral
2012;73:533-537.
sure therapy for anxiety disorders in humans? A meta-analysis. PLoS ONE.
2014;9:e93519.
62. Wilhelm S, Buhlmann U, Tolin DF et al. Augmentation of behavior therapy with
D-cycloserine for obsessive-compulsive disorder. Am J Psychiatry 2008;165:335-
341.
64. Andersson E, Hedman E, Enander J et al. D-cycloserine vs placebo as ad-
junct to cognitive behavioral therapy for obsessive-compulsive disorder and
interaction with antidepressants. A randomized clinical trial. JAMA psychiatry
2015;doi:10.1001/jamapsychiatry.2015.0546
65. Rothbaum BO, Price M, Jovanovic T et al. A Randomized, Double-Blind Evaluation
of D-Cycloserine or Alprazolam Combined With Virtual Reality Exposure Therapy
for Posttraumatic Stress Disorder in Iraq and Afghanistan War Veterans. Am J
Psychiatry 2014;171:640–648.
66. Smits JA, Rosenfield D, Davis ML et al. Yohimbine enhancement of exposure
therapy for social anxiety disorder: a randomized controlled trial. Biol Psychiatry
2014;75:840-846.
67. Morrison FG, Ressler KJ. From the neurobiology of extinction to improved clinical
treatments. Depress Anxiety 2014;31: 279-290.
68. de Bitencourt RM, Pamplona FA, Takahashi RN. A current overview of cannabi-
noids and glucocorticoids in facilitating extinction of aversive memories: potential
extinction enhancers. Neuropharmacology 2013;64:389-395.
122
123
124
Chapter 6
Nederlandse samenvatting
(Dutch summary)
125
Chapter 6
Angststoornissen zijn veel voorkomende psychiatrische aandoeningen (ongeveer 1 op

de 5 Nederlanders heeft, op enig moment in het leven een angststoornis). Onder de
angststoornissen vallen de paniekstoornis (met of zonder agorafobie), de sociale fobie,
de specifieke fobie, de agorafobie (zonder paniekstoornis), de gegeneraliseerde angst-
stoornis (GAS), de obsessief-compulsieve stoornis (OCD), de posttraumatische stress
stoornis (PTSS) en de acute stress stoornis. Complexe angststoornissen zijn angst-
stoornissen die vaak ernstig zijn, gepaard gaan met andere stoornissen (comorbiditeit)
en die moeilijk te behandelen zijn (therapie resistent). Aangezien complexe angst-
stoornissen dus vaak een ernstig en chronisch beloop hebben is onderzoek naar risico
factoren, onderliggende mechanismen en nieuwe behandelmogelijkheden gewenst. In
dit proefschrift worden deze drie aspecten belicht en onderzocht. In het eerste deel van
het proefschrift worden drie studies gepresenteerd waarin comorbiditeit (als mogelijke
risico factor voor complexe angststoornissen) wordt onderzocht. In het tweede deel
wordt een studie beschreven waarin een mogelijk onderliggend mechanisme dat een
rol speelt in de ontwikkeling van angststoornissen (de “expectancy bias”) wordt onder-
zocht en in het derde deel wordt een studie naar een nieuwe behandeloptie (d-cyclo-
serine) gepresenteerd. De drie onderdelen van dit proefschrift worden in dit hoofdstuk
apart samengevat, gevolgd door enkele algemene conclusies.
Risicofactor voor complexe angststoornis; comorbiditeit

Achtergrond
Comorbiditeit bij angststoornissen komt veel voor en is geassocieerd met ernstigere
symptomen, een slechter beloop en verminderd functioneren.
Omdat comorbiditeit zo’n belangrijk gegeven is in het chronisch beloop van angst-
stoornissen, zal hieronder eerst kort ingegaan worden op chroniciteit en vervolgens
een aantal aspecten van comorbiditeit verder worden toegelicht.
De interpretatie van onderzoek naar chroniciteit wordt bemoeilijkt door het ontbre-
ken van een eenduidige definitie van chroniciteit. Chroniciteit wordt gebruikt voor “het
voortbestaan van angstsymptomen die voldoen aan de criteria voor een angststoornis”,
maar ook voor “het voortbestaan van enige angstsymptomen” (zonder dat voldaan
wordt aan de criteria voor een stoornis). Daarnaast worden chroniciteit en therapie
resistentie vaak door elkaar gebruikt, hoewel dit overlappende, maar onderscheidbare
fenomenen zijn. Een chronisch beloop is gedefinieerd als “het voortbestaan van symp-
tomen, ongeacht de behandeling”, terwijl therapie resistentie wordt gedefinieerd als
“het voortbestaan van symptomen gedurende een adequate behandeling”.
Bij schizofrenie en depressie is chroniciteit goed onderzocht, in angststoornissen
veel minder. Dit gebrek aan onderzoek is goed zichtbaar in de huidige multidisciplinai-
re richtlijn; in de richtlijn voor angststoornissen komt de term “chronisch” (of “therapie
resistent”) slechts zeven keer voor, terwijl er een heel hoofdstuk gewijd is aan therapie
resistentie in de richtlijn schizofrenie en een aparte richtlijn is voor chronische depres-
sie (www.ggzrichtlijnen.nl).
Er is een aantal grote projecten opgezet om dit hiaat in kennis te dichten; the Har-
vard/Brown Anxiety Research Program (HARP), De Netherlands Study of Depression
and Anxiety (NESDA), de Netherlands Obsessive Compulsive Disorder Association
(NOCDA) studie en the Brown Longitudinal Obsessive Compulsive Study (Brown studie).
126
Nederlandse samenvatting (Dutch summary)
Deze studies hebben de kennis over het chronische beloop van angststoornissen aan-
zienlijk vergroot en ook de rol van comorbiditeit hierin meegenomen. Er is bijvoorbeeld
door dit onderzoek aangetoond dat comorbiditeit een belangrijke onafhankelijke voor-
speller van chroniciteit is, naast ernst van de ziekte. Het is van belang om te melden dat
deze grote onderzoeken niet alle angststoornissen hebben meegenomen. De HARP en
de NESDA studie hebben specifiek gekeken naar de paniekstoornis (met of zonder agor-
afobie), de sociale fobie en de gegeneraliseerde angststoornis. De NOCDA en de Brown
studie hebben patiënten onderzocht met een obsessief-compulsieve stoornis (OCD).
Een tweetal belangrijke aspecten van comorbiditeit moeten nog genoemd worden.
Ten eerste; de definitie van comorbiditeit. De officiële definitie van comorbiditeit is: “het
tegelijk voorkomen van twee afzonderlijke ziekten”. Dat vormt een probleem in de psy-
chiatrie, aangezien de huidige psychiatrische aandoeningen geclassificeerd worden op
basis van de DSM-IV: Deze is gebaseerd op symptoom clusters en niet op afzonderlij-
ke ziekten. Angst en depressie hebben, volgens de huidige DSM- classificatie, overlap-
pende symptomen. Daarnaast zijn ook de huidige behandelingen deels overlappend.
Critici van het DSM- classificatiesysteem noemen psychiatrische comorbiditeit dan ook
een kunstmatig bijproduct van dit systeem en zien comorbiditeit meer als een ernst-
maat: mensen met comorbide psychiatrische aandoeningen zijn er simpelweg slechter
aan toe. En hoewel deze kritiek (deels) gegrond is, is er tot nu toe geen betere manier
gevonden om psychiatrische ziekten te definiëren. Een indeling op bijvoorbeeld geneti-
sche of neurobiologische oorsprong bestaat (nog) niet.
Het tweede belangrijke aspect sluit hierbij aan: behandelaren blijken terughoudend
in het diagnosticeren van alle comorbide stoornissen. Mogelijk denken ze dat het niet
van belang is in de behandeling. Dat idee wordt gevoed door het gebrek aan aandacht
voor comorbiditeit in de huidige richtlijnen voor behandelaars.
Doel van het onderzoek

In het eerste deel van dit proefschrift worden drie studies naar comorbiditeit bij angst-
stoornissen gepresenteerd met als doel de kennis over dit onderwerp te vergroten om 6
uiteindelijk klinische implicaties en richtlijnen voor de behandeling van comorbiditeit te
bieden. Daarnaast is de verwachting dat behandelaren systematischer comorbiditeit
zullen diagnosticeren wanneer blijkt dat comorbiditeit belangrijk is voor de behan-
deling. In hoofdstuk 2.1 wordt de studie beschreven waarin werd onderzocht of het
klinisch relevant is om in angststoornissen onderscheid te maken in type comorbiditeit
en het aantal comorbide stoornissen. In hoofdstuk 2.2 wordt een studie beschreven
waarin het voorkomen van comorbiditeit bij patiënten met een obsessief-compulsieve
stoornis wordt onderzocht. Ook werd gekeken naar de invloed van comorbiditeit op het
beloop van OCD. In hoofdstuk 2.3 is de impact van comorbide obsessieve compulsieve
symptomen (OC-symptomen, maar geen volledige obsessief-compulsieve stoornis) bij
patiënten met angst- en depressieve stoornissen onderzocht.
Bevindingen
In hoofdstuk 2.1 wordt de studie gepresenteerd waarin 1004 patiënten zijn geïnclu-
deerd uit het NESDA onderzoek. Deze patiënten hadden allemaal een angststoornis
en zijn in verschillende groepen verdeeld en onderling vergeleken; patiënten met een
127
Chapter 6
pure angststoornis, met angst-angst comorbiditeit, met angst-depressie comorbiditeit

en met ‘dubbele’ comorbiditeit (zowel een comorbide andere angststoornis als een
comorbide depressieve stoornis). Comorbiditeit bij angststoornissen hing samen met
meer klachten en meer chroniciteit ten opzichte van de pure angststoornissen. Bo-
vendien werd aangetoond dat angst-angst comorbiditeit een ander beloop had dan
comorbiditeit tussen angst en depressie waarbij er bij angst-angst comorbiditeit meer
beperkingen in het dagelijks leven werden gerapporteerd. Er werd geconcludeerd dat
in de klinische praktijk alle comorbide stoornissen moeten worden gediagnosticeerd,
gezien de hoge mate van ernst, het ongunstige beloop en de grote beperkingen in het
dagelijks leven. Patiënten met een angststoornis en comorbiditeit moeten een snelle en
adequate behandeling krijgen.
In hoofdstuk 2.2 wordt een onderzoek beschreven waarin 382 patiënten met OCD
uit het NOCDA onderzoek werden onderzocht. Comorbiditeit, in het bijzonder van meer-
dere comorbide stoornissen, bleek samen te hangen met meer chroniciteit, ernstige
OCD en meer beperkingen in het dagelijks leven. Er werd dan ook aanbevolen dat er
rekening gehouden moet worden met comorbiditeit bij de planning van de behandeling.
In hoofdstuk 2.3 wordt een studie gepubliceerd waarin 2125 patiënten uit de NES-
DA studie zijn opgenomen; gezonde proefpersonen, patiënten met een angst of de-
pressieve stoornis in de voorgeschiedenis en patiënten met een huidige angst en/of
depressieve stoornis. Klinisch relevante OC -symptomen bleken veel voor te komen bij
patiënten met een angst en/of depressieve stoornis. Bovendien zijn klinisch relevante
OC-symptomen geassocieerd met een hogere ernst en een ongunstig beloop. Er werd
aanbevolen dat clinici alert zijn op OC-symptomen bij patiënten met een angst of de-
pressieve stoornis, omdat dit een ongunstig beloop voorspelt. Bovendien moet worden
overwogen de aanwezige OC-symptomen te behandelen, ook als er niet aan de volle-
dige criteria voor een OCD wordt voldaan.
Conclusies
In alle drie de studies worden aanbevelingen gedaan om comorbiditeit tijdig en adequaat
te behandelen. We bevelen dan ook aan om comorbiditeit direct te diagnosticeren en niet
pas in stap drie van de beslisboom mee te nemen, zoals de huidige richtlijn suggereert.
Maar wat is een adequate behandeling van comorbiditeit? Die vraag blijft nog gro-
tendeels onbeantwoord, hoewel er wel op basis van de literatuur over dit onderwerp
een aantal opmerkingen en aanbevelingen te maken is.
Ten eerste; het combineren van medicatie en psychotherapie lijkt een goede optie
om ook comorbiditeit te behandelen, echter literatuur over dit onderwerp blijft beperkt
tot angststoornissen met een comorbide ernstige depressie. De meeste farmacologi-
sche behandelstudies bij angststoornissen sluiten patiënten met comorbiditeit uit.
Ten tweede, het tegelijkertijd aanbieden van twee psychotherapeutische behande-
lingen is niet effectiever gebleken dan één behandeling. De beslissing om één stoornis
goed te behandelen geeft soms ook al een vermindering van de comorbide stoornis.
Echter, het is belangrijk om te melden dat een onvolledig herstel van de comorbide
stoornis een grotere kans op terugval geeft.
Ten derde, een nieuwe psychotherapeutische optie is de zogenaamde “transdi-
agnostische” behandeling; een groepsbehandeling die niet gericht is op één stoornis,
128
maar op algemene psychotherapeutische elementen bijvoorbeeld emotie regulatie of

exposure. Deze groepsbehandelingen lijken even effectief als groepsbehandelingen
die zich op één specifieke stoornis richten en zouden mogelijk ook comorbide stoornis-
sen kunnen behandelen. Deze veronderstelling moet nog verder onderzocht worden.
Onderliggend mechanisme van complexe angststoornissen; expectancy bias.

Achtergrond
In hoofdstuk drie is een psychologisch mechanisme onderzocht dat mogelijk een rol
speelt bij het ontwikkelen en voortbestaan van angststoornissen: de “expectancy bias”.
Dit is het beste te vertalen als “de neiging om gevaar te overschatten”. Mensen met
een angststoornis hebben de neiging om de kans op een negatieve gebeurtenis na een
angst-gerelateerde of neutrale gebeurtenis hoger in te schatten dan mensen zonder
angststoornis. Een voorbeeld: patiënten wordt gezegd de kans in te schatten dat ze
een schok krijgen na het zien van een bepaald plaatje; patiënten met een angststoornis
schatten de kans op een schok hoger dan mensen zonder angststoornis; zowel na het
zien van angst-gerelateerde plaatjes als na neutrale plaatjes. Dit mechanisme is mogelijk
(deels) de oorzaak van het ontstaan of van het chronische beloop van angststoornissen.

In dit hoofdstuk wordt een studie gepresenteerd waarin werd onderzocht of patiënten
met een paniekstoornis met agorafobie een “expectancy bias” hebben. Ook werd on-
derzocht of een hogere “expectancy bias” samenhangt met het effect van exposure
therapie bij patiënten met een paniekstoornis en agorafobie. Tijdens exposure therapie
worden patiënten blootgesteld aan hun angst en gaan ze, stapsgewijs, de confrontatie
met hun angst aan.
Bevindingen
Er werd een “expectancy bias” taak gebruikt waarin deelnemers (70 patiënten en 65
gezonde proefpersonen) blootgesteld werden aan een reeks van paniek-gerelateerde 6
en neutrale woorden. Aan hen werd gevraagd of ze verwachtten dat het woord zou
worden gevolgd door een hard geluid. Patiënten met een paniekstoornis bleken inder-
daad een grotere verwachting aan te geven om het harde geluid te horen na paniek-ge-
relateerde woorden, vergeleken met neutrale woorden. Bovendien was de verwachting
om het harde geluid te horen bij patiënten hoger dan bij de gezonde proefpersonen.
Een hogere “expectancy bias” bleek niet samen te hangen met de uitkomst van ex-
posure therapie. Met andere woorden; er was bij patiënten wel een grotere neiging om
gevaar te overschatten, maar deze “expectancy bias” bleek geen voorspeller van een
slecht resultaat op exposure therapie.
Conclusies
Dit deel van het proefschrift bevat een meer experimenteel onderzoek, echter enkele
klinische implicaties kunnen worden vermeld. Ten eerste moeten behandelaren zich
bewust zijn van de aanwezigheid van een “expectancy bias” bij patiënten met een pa-
niekstoornis met agorafobie. Deze neiging om dreiging te overschatten moet uitgelegd
worden en onderwerp in de therapie zijn. Ten tweede, een algemenere conclusie; het
129
Chapter 6
is van belang om experimenteel onderzoek uit te voeren om zo de verschillende me-

chanismen die een rol spelen bij het ontwikkelen en het voortbestaan van angststoor-
nissen te doorgronden. Dit kan leiden tot betere behandeling opties.
Nieuwe behandeloptie bij complexe angststoornissen; d-cycloserine

Achtergrond
De huidige behandeling van angststoornissen bestaat uit medicatie en/of psychothera-
pie (bijvoorbeeld exposure therapie en/of cognitieve gedragstherapie). De combinatie
van medicatie en therapie heeft tot nu toe maar een beperkte meerwaarde laten zien
boven het gebruik van één van deze twee. Bovendien zijn er veel patiënten die niet (of
niet volledig) reageren of die terugvallen ondanks een behandeling. De laatste jaren
zijn er nieuwe ontwikkelingen met betrekking tot het combineren van medicatie en the-
rapie. Een interessante ontwikkeling is die van d-cycloserine (DCS); DCS werkt via de
beïnvloeding van de N-methyl-D-aspartaat (NMDA) receptor, één van de receptoren
van het glutamaat-systeem die een belangrijke rol speelt bij het aanleren, beklijven en
afleren van angst. DCS zou dus het effect van de therapie kunnen versterken door het
aan- en afleren van angst te verbeteren. DCS is een middel dat slechts enkele keren
ingenomen wordt tijdens de therapie. Daarmee werkt het anders dan de medicatie die
nu gebruikt wordt; die moet dagelijks en vaak jarenlang ingenomen worden.
In onderzoek met dieren is aangetoond dat DCS het effect van therapie versterkt.
Ook in studies met patiënten is dit inmiddels aangetoond; patiënten die DCS innemen
voor een therapie sessie, lijken meer baat te hebben bij deze sessie; de angstklachten
nemen af en ze herstellen eerder. De precieze dosering, het tijdstip van dosering en
de frequentie van DCS inname is echter nog niet onderzocht. Daarnaast is gebleken
uit dieronderzoek dat DCS ook slechte therapie sessies kan versterken (door angst te
versterken). Tot slot is nog van belang te melden dat uit dieronderzoek is gebleken dat
DCS ook effectief is als het gegeven wordt ná de sessies. Dit komt waarschijnlijk om-
dat het beklijven van de therapie; het ‘leren’ ook in de uren na de therapie nog plaats
vindt. Het is van groot belang dit ook bij patiënten te onderzoeken. Mocht DCS ook
effectief zijn wanneer het na een sessie gegeven wordt, dan heeft dat veel klinische
implicaties. Zo zou DCS dan bijvoorbeeld in de toekomst alleen ná een succesvolle
sessie gegeven kunnen worden.

In hoofdstuk vier wordt een gerandomiseerd, placebo gecontroleerde studie gepresen-
teerd. Het effect van DCS werd onderzocht in een groep van 57 patiënten met paniek-
stoornis met agorafobie. Daarnaast is gekeken of DCS ook effectief is wanneer het na
de exposure sessies wordt gegeven. Alle patiënten kregen 12 sessies exposure thera-
pie (waarin ze stapsgewijs blootgesteld worden aan hun angst). Zes van de 12 sessies
werd gecombineerd met DCS. De patiënten zijn in drie groepen verdeeld; patiënten die
125 mg DCS een half uur voor de sessies innamen; patiënten die 125 mg DCS direct
na afloop van de sessies innamen en patiënten die placebo pillen kregen.
Bevindingen
Er werd geen verschil gevonden in effect van de therapie tussen patiënten die DCS of
130
placebo kregen. In onze studie leek er dus geen meerwaarde te zijn van het geven van
DCS bij exposure therapie; zowel inname van DCS vóór als ná de sessies gaf geen
extra verbetering vergeleken met de placebo groep. Wel waren er aanwijzingen dat
DCS ná de sessies effectiever is als DCS voor de sessies, maar het belang van deze
bevinding is twijfelachtig, omdat zowel DCS ná als DCS vóór de sessies niet effectiever
bleek dan placebo.
Als mogelijke oorzaken voor de negatieve bevindingen kan een aantal methodolo-
gische beperkingen worden genoemd; de groep (57 patiënten) is wel de grootste groep
tot nu toe onderzocht met DCS, maar de groep is onderverdeeld in drie groepen (DCS
voor exposure, DCS na exposure en placebo). Dit maakt dat de statistische power
mogelijk toch te klein is geweest om significante effecten aan te tonen. Daarnaast is
125 mg DCS gebruikt, terwijl een recent groot onderzoek aantoont dat een lage dose-
ring (50 mg) effectiever is dan een hoge dosering (500 mg). De weliswaar relatief lage
dosering van 125 mg in onze studie zou dus een suboptimale dosering kunnen zijn.
Echter, eerdere studies met 125 mg DCS vonden wel een effect. Bovendien, zolang er
geen duidelijkheid is over de optimale timing, dosering en frequentie van DCS, blijven
deze factoren speculatief. Tot slot zou het gelijktijdig gebruik van andere medicatie (an-
tidepressiva of benzodiazepines) onze resultaten negatief beïnvloed kunnen hebben.
Er is in één studie aangetoond dat DCS geen effect zou hebben als het gecombineerd
wordt met antidepressiva, echter een andere studie vond dit niet. In ander onderzoek
werd aangetoond dat benzodiazepines het effect van exposure therapie kunnen ver-
minderen. Om te kijken of het gelijktijdig gebruik van medicatie onze resultaten heeft
beïnvloed, is in een subgroep van medicatie naïeve patiënten (n=37) de analyses her-
haald. Ook in deze analyses werd geen effect van DCS aangetoond.
Conclusies
Concluderend zijn de resultaten van DCS als toevoeging aan therapie nog onvoldoen-
de eenduidig om DCS al te gebruiken in de dagelijkse praktijk. Echter, gegevens uit
de grote onderzoeken en de voorzichtige bevindingen in de gepresenteerde studie dat 6
DCS ná de sessies minstens even effectief is als ervoor, maken DCS een middel dat
zeker verder onderzocht moet worden.
Slotopmerkingen
Dit proefschrift bevat onderzoek naar drie belangrijke aspecten van complexe angst-
stoornissen; comorbiditeit als een risicofactor, “expectancy bias” als mogelijk onderlig-
gend mechanisme en DCS als veelbelovende nieuwe behandeloptie. Dit proefschrift
draagt bij aan meer kennis over complexe angststoornissen en aan een betere behan-
deling. Er is aangetoond dat comorbiditeit serieus genomen moet worden en dat comor-
biditeit vroeg in de behandeling meegenomen dient te worden in de behandelkeuze. Er
kan geconcludeerd worden dat de neiging om gevaar te overschatten groter is bij pati-
ënten met een paniekstoornis dan bij gezonde proefpersonen en dat dit een betrouw-
bare voorspeller van de respons op de behandeling zou kunnen zijn. En hoewel DCS
nog niet gebruikt kan worden in de dagelijkse praktijk, is het wel een middel dat verder
onderzoek verdiend omdat het mogelijk het effect van exposure therapie versterkt.
131
132
dankwoord
133
Dankwoord
Dan rest nog een dankwoord, het laatste, maar ook het leukste en makkelijkste deel
om te schrijven. Er zijn veel mensen die een aandeel hebben gehad in dit proefschrift
en het onderzoek dat erin beschreven staat. Een aantal mensen wil ik hier specifiek
bedanken.
Allereerst wil ik mijn promotoren en co-promotoren bedanken: Marcel van den
Hout, Ton van Balkom, Danielle Cath en Harold van Megen; Bedankt voor jullie tijd,
inzet en steun. Vier meningen bleken soms verdeeld te zijn, maar ik was blij met jullie
expertise, adviezen en kritiek. Marcel; jij was altijd betrokken en beschikbaar, in het
begin van een afstandje, naarmate het einde naderde steeds dichterbij. Ik kon altijd bij
je aankloppen en je hebt me vaak de juiste richting op gewezen. Jouw snelle reacties
met een altijd scherpe blik hebben de artikelen verdiept en verbeterd. Bedankt voor de
prettige samenwerking.
Ton; onze samenwerking was intensief in het begin en minder intensief aan het
eind; zoals een goede opleiding behoort te lopen; je liet me steeds zelfstandiger wer-
ken. Bedankt dat je me de mogelijkheid hebt gegeven om in mijn keuzejaar onderzoek
te combineren met het werken in de crisisdienst. Zo kon mijn behoefte aan actie ruim-
schoots gestild worden! Ik vond onze samenwerking erg prettig!
Danielle; ik herinner me onz eerste kennismaking nog heel goed; met Harold kwam
ik bij jou om te kijken of we samen aan het d-cycloserine onderzoek zouden kunnen
gaan beginnen. Jij had al een goed begin gemaakt met de subsidi-aanvraag en het
protocol, ik kon daar gemakkelijk instromen. Makkelijk, dat dachten we….. tot we naar
de METC moesten! Maar wat een leuk en gevarieerd project was het om te doen. Ik
ben blij met de mooie ervaringen en kansen die ik heb mogen krijgen.
Harold; jij hebt me aangenomen als AIOS, nog voordat mijn co-schappen erop za-
ten, je hebt me geïnspireerd om onderzoek te gaan doen en me alle kansen gegeven
me te ontwikkelen op alle gebieden. Zowel gedurende dit promotietraject ), als tijdens
mijn opleiding en ook nu, heb je me altijd gesteund. Heel fijn dat ik altijd bij je kan aan-
kloppen voor jouw mening; daar heb ik veel aan. Jouw gevoel voor humor, jouw scher-
pe blik, jouw creativiteit; ik hoop nog lang met je samen te werken.
Co-auteurs: Neeltje Batelaan, Puck Duits, Marleen Rijkeboer en Adriaan Hoogen-
doorn. Ik wil jullie bedanken voor de prettige en gezellige samenwerking. Jullie bij-
dragen heb ik altijd erg gewaardeerd en broodnodig gehad. Neeltje; jij begeleidde me
tijdens mijn eerste artikel en hebt me de basisprincipes van SPSS geleerd. Puck; mijn
d-cycloserine maatje. Wat was het fijn om met jou te kletsen, discussiëren en samen
te werken aan dit onderzoek en het schrijven van de artikelen. Marleen; jij hebt me de
basis van exposure therapie bijgebracht en het therapie protocol geschreven. Het was
ontzettend leuk en leerzaam voor mij om een spoed cursus te krijgen van zo’n expert
op dit gebied! Adriaan; werd de statistiek me te ingewikkeld, dan noemde Ton steevast
jouw naam. En terecht, jouw kritische blik, meedenken en ingewikkelde syntaxen heb-
ben de onderzoeken solide en betrouwbaar gemaakt.
Onderzoeksgroep: Aart de Leeuw, Henny Visser, Harold van Megen en Hanneke
du Mortier. Dank voor de fijne samenwerking, die hopelijk nog lang zal voortduren. Ik
heb het als een mooie kans en heel inspirerend ervaren om in de GGZ zo’n productie-
ve, prettige, enthousiaste onderzoeksgroep tegen te komen. Henny: ik weet nog goed
134
Dankwoord
dat wij eens samen naar huis reden en ik erg stellig zei dat ik niet snap waarom som-
mige mensen alles tegelijk wilden doen; opleiding, gezin, promoveren… en moet je mij
nu eens zien!
Intervisie genoten: Giorgio, Annemiek, Nienke, Iris, Annelies, Michiel: wat hebben
we al veel meegemaakt; begonnen als AIOS en inmiddels allemaal psychiater. Dat er
nog maar heel veel gezellige etentjes/intervisies mogen volgen! Andre, Gerco, Irma,
Isis, Annette, Diewertje, Marscha: Ik ben de tweede in ons groepje die promoveert,
nog twee te gaan…. En een speciale dank aan Annemiek; toen ik begon met het on-
derzoek hoopte ik al dat jij mijn paranimf zou worden, mocht een promotie er ooit van
komen. Jij hebt me immers aangenomen! Onze samenwerking was vanaf de eerste
dag zo soepel en natuurlijk. Je was er altijd om te overleggen of om gewoon gezellig te
kletsen. Wat jammer dat we nu (even) niet meer samenwerken!
Collega’s en oud collega’s: bedankt voor jullie steun tijdens dit project, met name
dank aan ‘het Johannesbos team’; verpleging, Hetty, Zegert, Roy en Yolande. Jullie
hebben het voor mij in mijn eerste baan heel makkelijk gemaakt. Ik vond het een voor-
recht om het werken op de gesloten afdeling te combineren met het afronden van mijn
proefschrift.
Patiënten, onderzoeksassistenten, datamanagers en therapeuten: Dank voor jullie
inzet en deelname. Zonder jullie was er geen onderzoek van de grond gekomen.
Familie en vrienden: Dank voor het aanhoren van mijn verhalen, voor het ‘er zijn’
en voor alle leuke dingen die we hebben meegemaakt en hopelijk nog meemaken; wie
had dat gedacht, toen we in 1998(!) bij de Kabouterhut kwamen werken! Inmiddels bij-
na 20 jaar geleden en nog steeds de vaste ‘gang’. We zeggen het niet vaak, maar laat
ik het nu maar eens doen; ik voel me erg gelukkig met zo’n groep ‘vrienden voor het
leven’. Noëlla, lief oudste vriendinnetje: Ik ben erg blij met jou! Met jouw altijd nuchtere
kijk, jouw oog voor (grammatica-) details en jouw relativeringsvermogen (‘gaat je pro-
motie je nog meer tijd kosten in de toekomst dan dat je er tot nu toe in hebt gestoken?
Nee? Dan maak je het dus gewoon af!’).
Lieve pa en ma, lieve zusjes: Jullie maken mijn leven zoveel makkelijker. Onvoor-
waardelijke liefde en steun; jullie zijn er altijd voor me, ik hou van jullie.
Lieve Thomas en Sanne; jullie zijn te jong om erbij te zijn en te jong om te beseffen
wat jullie aandeel is geweest in dit proefschrift. Ik dacht altijd dat het zwaar zou zijn
om een gezin te combineren met promoveren, maar niks is minder waar. Het is zoveel
makkelijker te werken als je weet hoe leuk het is om thuis te komen! Het is zoveel mak-
kelijker om te schrijven, als je weet dat er thuis zoveel liefde wacht!
Tot slot; lieve Martijn…. Toch een plekje in mijn dankwoord veroverd! We hebben er
vaak grapjes over gemaakt, maar jouw onvoorwaardelijke vertrouwen in en steun aan
dit project heeft er uiteindelijk toe geleid dat dit boekje er nú ligt. Bedankt voor jouw ge-
duld, het aanhoren van mijn eindeloze ‘gezeur’ en bovenal; bedankt dat je er bent.
135
136
vitae
curriculumie
publicat s n
voordrachte
137
Curriculum vitae
Curriculum Vitae
Mieke Klein Hofmeijer-Sevink werd geboren in 1980 in Gorssel, groeide op in Slochte-
ren en kwam op haar 10e naar Leusden. Na haar eindexamen (1998) ging ze een jaar
naar Londen als au-pair om vervolgens Geneeskunde te studeren aan de Universiteit
van Amsterdam. In 2006 begon zij aan de opleiding tot psychiater bij het toenmalige
Meerkanten en combineerde dat met een promotie onderzoek. Van 2011-2015 werkte
ze op de gesloten opname afdeling in Ermelo, daarna op de polikliniek stemming en
angst van Symforameander in het Meander MC te Amersfoort. Daarnaast coördineert
zij het co-schap psychiatrie voor GGZ Centraal.
138
Publicaties, voordrachten
Publicaties
M. Klein Hofmeijer-Sevink, Z.J. van Eijk & R. van der Zwaard (2009). Acquired stutte-
ring in a psychotic patient. Tijdschrift voor Psychiatrie 51, 53 – 57.
A.S. de Leeuw, M. Klein Hofmeijer-Sevink, P. Duits & D. Cath. ‘D-cycloserine bij

angststoornissen’ in: Handboek angststoornissen. A.J.C.M. van Balkom, D. Ooster-
baan, I. van Vliet en S. Visser (Red). 2010. In druk.
M. Klein Hofmeijer-Sevink, H.G.M. van Megen en D. Cath (2010). D-cycloserine bij

angststoornissen. Silhouet, voorjaar 2010, pag 8-11.
Klein Hofmeijer-Sevink M, Batelaan NM, van Megen HJ, Penninx BW, Cath DC, van
den Hout MA, van Balkom AJ. (2012). Clinical relevance of comorbidity in anxiety disor-
ders: A report from the Netherlands Study of Depression and Anxiety (NESDA). Journal
of Affective Disorders 2012; 137(1-3):106-12.
Klein Hofmeijer-Sevink M, van Oppen P, van Megen HJ, Batelaan NM, Cath DC, van
der Wee NJ, van den Hout MA, van Balkom AJ. (2013). Clinical relevance of comorbidi-
ty in obsessive compulsive disorder: The Netherlands OCD Association study. J Affect
Disord. doi:pii: S0165-0327(13)00239-5.
M. Klein Hofmeijer-Sevink (2004). Symptomatology in a prodromal group and a healt-

hy control group, with and without cannabis use. Scriptie in het kader van de weten-
schappelijke stage.
M. Klein Hofmeijer-Sevink, M.M. Rijkeboer, D.C. Cath & P. Duits (2010). Protocol ex-
posure therapie bij paniekstoornis. Protocol in het kader van het onderzoek ‘ D-cyclo-
serine augmentatie van exposure therapie’.
Voordrachten
Voorjaarscongres Nederlandse Vereniging voor Pyschiatrie 2011:
Symposium lecture ‘D-cycloserine in anxiety disorders’.

Symposium lecture ‘comorbidity in OCD’.
Najaarscongres vGCt 2011:

Symposium ‘D-cycloserine’.

Symposium lecture ‘D-cycloserine bij angststoornissen’.
Voorjaarsconges Nederlandse Vereniging voor Pyschiatrie 2013:

Sympsoium lecture ‘comorbiditeit bij OCD. Resultaten van de NOCDA studie’.
139
Voordrachten
NEDKAD congres 2013:

‘’comorbiditeitspatronen bij OCD, resultaten van de NOCDA studie’’.
Voorjaarsconges Nederlandse Vereniging voor Pyschiatrie 2014:

Symposium lecture ‘comorbiditeit bij OCD. Resultaten van de NOCDA studie’.
140

KleinHofmeijer PDF

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

KleinHofmeijer PDF

Enviado por

Direitos autorais:

Formatos disponíveis

m ple x an x ie ty

Mieke Klein Hofmeijer-Sevink

Cover design: Ruben Boerhof (www.yellowlemmontree.nl)

(met een samenvatting in het Nederlands)

Mieke Klein Hofmeijer-Sevink

geboren op 7 mei 1980 te Gorssel

Chapter 1 General introduction 6

Chapter 2.1 Clinical relevance of comorbidity in anxiety 24

Chapter 2.2 Clinical relevance of comorbidity in Obsessive 40

Chapter 2.3 Occurrence and predictive value of clinically 60

Chapter 3 Expectancy bias and treatment outcome in patients 76

Chapter 4 D-cycloserine does not enhance exposure therapy 90

Chapter 5 Summary and general discussion 108

Chapter 6 Nederlandse samenvatting (Dutch summary) 124

Curriculum vitae, publicaties, voordrachten 136

Chronicity in anxiety disorders

OCD needs special attention in this introduction. Although categorised as an

Risk factor for complex anxiety disorders; comorbidity

Clinical relevance of comorbidity

Comorbidity in the light of the DSM

Underlying mechanism of complex anxiety disorders; expectancy bias

Fear acquisition and extinction

Problems in fear acquisition and extinction

Treatment enhancement in anxiety disorders; d-cycloserine

studies found DCS augmentation post extinction learning experiments to be equally

Aims of this thesis

Finally, in Chapter 5, a summary of the main results is provided and discussed in a

17. Merikangas KE, Zhang H, Avenevoli S et al. Longitudinal Trajectories of Depres-

conditioned freezing. Behav. Neurosci 2003;117:341-349.

to psychotherapy: use of D-cycloserine in phobic individuals to facilitate extinction

Clinical relevance of comorbidity in anxiety disorders:

Mieke Klein Hofmeijer-Sevink, Neeltje M. Batelaan, Harold J.G.M. van Megen,

Journal of Affective Disorders 2012;137:106-112.

Aims of the study

1) anxiety-anxiety comorbidity differs from a single anxiety disorder

Vulnerability factors included a family history of anxiety and depressive disorders,

Female gender 72.1 (724)

Partner 66.7 (670)

Chronicity 45.3 (455)

Treatment 51.3 (515)

Age (years) 41.0 (12.6)

Education (years) 11.7 (3.2)

Age of onset (years) 20.9 (12.3)

Total IDS score 28.3 (12.6)

Total BAI score 17.7 (10.7)

Total FQ score 34.7 (20.1)

Cognitive disability 31.5 (20.9)

Physical disability 18.7 (21.7)

Social disability 32.8 (20.6)

comorbidity (n=105). The results are presented in Tables 2 and 3.

Group (total n=1004) comparison

Group (total n=1004) comparison

Number of comorbid disorders

Sociodemographics and vulnerability factors of comorbidity

Limitations and strengths

primary care setting. J Clin Epidemiol 2003;56:507-514.

Clinical relevance of comorbidity in Obsessive Compulsive

Journal of Affective Disorders 2013; 150:847-854.

were considered significant. Subsequently, bivariate and multivariate logistic regres-

Comparison of pure OCD to OCD with current comorbidity

Diagnosis % (n) % (n)

Pure OCD 45.0 (172) 22.3 (85)

OCD + any Anxiety Disorder 36.6 (140) 46.1 (176)