Page 1
Linköpings university | Department of Behavioural Sciences and Learning
The Psychologist Programme
Spring semester of 2016
Prediction of treatment response
in Social Anxiety Disorder -
what does the brain tell us that
questionnaires do not?
– Using brain activity related to self- and other-referential
criticism to predict treatment response to Internet-
delivered Cognitive Behaviour Therapy for Social
Anxiety Disorder
Nils Isacsson
Örn Kolbeinsson
Linköpings universitet
SE-581 83 Linköping, Sweden
013-28 10 00, www.liu.se
Page 2
The Psychologist Programme consists of 300 academic credits taken over the
course of five years. The programme has been offered at Linköping University
since 1995. The curriculum is designed so that the studies focus on applied
psychology and its problems and possibilities from the very beginning. The
coursework is meant to be as similar to the work situation of a practicing
psychologist as possible. The programme includes one placement period, totaling
12 weeks of full time practice as well as clinical practice at the programmes own
clinic. Studies are based upon Problem Based Learning (PBL) and are organized in
eight themes, Introduction 7,5 credits, Cognitive psychology and the biological
bases of behavior, 37,5 credits; Developmental and educational psychology, 52,5
credits; Society, organizational and group psychology, 60 credits; Personality
theory and psychotherapy, 67,5 credits; Placement practice and regarding the
Psychologist as profession, 27,5 credits; Research methods, 17,5 credits and degree
paper 30 credits.
This report is a psychology degree paper, worth 30 credits, spring semester 2016.
Main supervisor Gerhard Andersson and associate supervisor Kristoffer NT
Månsson.
Department of Behavioral Sciences and Learning
Linköping University
581 83 Linköping
Telephone +46 (0)13-28 10 00
Fax +46 (0)13-28 21 45
Page 3
Division, Department
Department of Behavioural Sciences and Learning
581 83 Linköping
SWEDEN
Date
2016-05-26
Language Report category ISRN-nummer LIU-IBL/PY-D—16/422—SE
Svenska/Swedish
X Engelska/English
Licentiate dissertation
Degree project
Bachelor thesis
X Master thesis
Other report
Title Prediction of treatment response in Social Anxiety Disorder, what does the brain tell us that
questionnaires do not? - Using brain activity related to self- and other-referential criticism to predict
treatment response to Internet-delivered Cognitive Behavioural Therapy for Social Anxiety Disorder
Authors Nils Isacsson and Örn Kolbeinsson
Abstract
Predicting who will benefit from what in the treatment of psychiatric disorders is incremental to future
development of psychological treatments. In the current study functional magnetic resonance imaging
(fMRI) data from participants with social anxiety disorder (SAD) was used to elucidate whether neural
responses to negative evaluation could predict treatment response in SAD. Nine weeks prior to Internet-
delivered Cognitive Behaviour Therapy (ICBT) onset, participants viewed negative social stimuli directed
either at themselves or an significant other during fMRI scanning. Regression analyses including the
differential activations for other-referential criticism in contrast to self-referential criticism in the posterior
mid cingulate cortex (pMCC) and the lingual gyrus (LG) predicted 34% of treatment change as measured by
residual gain scores on the Liebowitz Social Anxiety Scale Self-Report (LSAS-SR) in our sample. The final
regression model, combining these measures with behavioural measures, which by themselves explained
27% of the variance, resulted in a model explaining 50% of the variance regarding treatment response. This
lends additional support to the notion that further elucidating the neurobiological underpinnings of core
processes in SAD, as well as the neural correlates of treatment response to CBT, would be of great value in
predicting treatment outcome.
Keywords
Social Anxiety Disorder, Prediction, Neuroimaging, Functional Magnetic Resonance Imaging, Cingulate
cortex, Lingual Gyrus, Internet-delivered Cognitive Behaviour Therapy.
Page 4
Abstract Predicting who will benefit from what in the treatment of psychiatric disorders is
incremental to future development of psychological treatments. In the current study
functional magnetic resonance imaging (fMRI) data from participants with social
anxiety disorder (SAD) was used to elucidate whether neural responses to negative
evaluation could predict treatment response in SAD. Nine weeks prior to Internet-
delivered Cognitive Behaviour Therapy (ICBT) onset, participants viewed negative
social stimuli directed either at themselves or a significant other during fMRI
scanning. Regression analyses including the differential activations for other-
referential criticism in contrast to self-referential criticism in the posterior mid
cingulate cortex (pMCC) and the lingual gyrus (LG) predicted 34% of treatment
change as measured by residual gain scores on the Liebowitz Social Anxiety Scale
Self-Report (LSAS-SR) in our sample. The final regression model, combining
these measures with behavioural measures, which by themselves explained 27% of
the variance, resulted in a model explaining 50% of the variance regarding
treatment response. This lends additional support to the notion that further
elucidating the neurobiological underpinnings of core processes in SAD, as well as
the neural correlates of treatment response to CBT, would be of great value in
predicting treatment outcome.
Page 5
Acknowledgements We would like to extend a warm thank you to Kristoffer N.T. Månsson for giving
us this opportunity to delve deeper than we would have ever imagined into the
inner workings of anxiety, in an MRI-lab! Without you none of this would have
been possible, thank you. We would also like to thank Gerhard Andersson for his
encouraging words and constructive feedback.
Our gratitude also goes out to everyone involved in the project, Carl-Johan
Boraxbekk, Tomas Furmark, Håkan Fischer, the lab-staff at Umeå center for
Functional Brain Imaging, and last but not least the participants.
Page 6
Table of Contents Background ................................................................................................................................................. 1
Social Anxiety Disorder ............................................................................................................................ 1
Cognitive Behaviour Therapy for Social Anxiety Disorder ..................................................................... 2
Neuroimaging ........................................................................................................................................... 4
Neurobiology of Social Anxiety Disorder ............................................................................................ 5
Self-related processing and neural-correlates ....................................................................................... 9
Prediction of treatment response ............................................................................................................. 10
Depression ........................................................................................................................................... 10
Pretreatment symptom severity of Social Anxiety Disorder ............................................................... 12
Treatment expectancy ......................................................................................................................... 12
Biological predictors of treatment response ........................................................................................ 13
Aims ............................................................................................................................................................ 15
Materials and methods ............................................................................................................................. 15
General procedure ................................................................................................................................... 15
Procedure ................................................................................................................................................ 15
Participants .............................................................................................................................................. 17
Behavioural measures ............................................................................................................................. 18
Treatment ................................................................................................................................................ 19
Experimental task .................................................................................................................................... 19
Data acquisition and preprocessing......................................................................................................... 20
Data analysis ........................................................................................................................................... 21
Behavioural measures predicting outcome ......................................................................................... 22
Analysis of imaging data..................................................................................................................... 22
Results ........................................................................................................................................................ 24
Exploratory analyses of activation related to self- and other-referential criticism ................................. 24
Predicting treatment response using behavioural measures and demographic characteristics ............... 25
Prediction analyses using brain responses .............................................................................................. 26
Discussion .................................................................................................................................................. 30
Differential responses to self- and other-referential criticism ............................................................. 30
Predicting treatment response using behavioural measures and demographic characteristics ........... 32
Prediction analyses correlating brain responses to residual gain scores ............................................. 34
Limitations .............................................................................................................................................. 37
Page 7
Future research ........................................................................................................................................ 38
Conclusions ............................................................................................................................................. 39
References .................................................................................................................................................. 40
Appendix A ................................................................................................................................................ 65
Appendix B ................................................................................................................................................ 66
Appendix C ................................................................................................................................................ 69
Appendix D ................................................................................................................................................ 70
Appendix E ................................................................................................................................................ 71
Appendix F ................................................................................................................................................ 72
Appendix G ................................................................................................................................................ 73
Appendix H ................................................................................................................................................ 74
Page 8
1
Background
Social Anxiety Disorder
Anxiety is a debilitating state defined by fear, worry, and stress. When these
reactions become so severe that they hinder the daily life of an individual and
cause suffering it is usually defined as an anxiety disorder (American Psychiatric
Association [APA], 2013b). The lifetime prevalence of any anxiety disorder has
been estimated to between 28.9% to 49.5% (Kessler, et al., 2005; Moffitt et al.,
2010), the estimations however vary depending on methodology, country of origin
and diagnostic manuals (Somers, Goldner, Waraich & Hsu, 2006; Moffitt et al.,
2010). Estimates from previously mentioned studies seemingly correspond to
recent survey estimates in the Swedish population (Statistics Sweden, 2015).
In prevalence studies of anxiety disorders Social Anxiety Disorder (SAD), first
described in the early 1900’s as social phobia (Janet, 1903), is the second most
common anxiety disorder, closely trailing specific phobia (Fehm, Pelissolo,
Furmark & Wittchen, 2005; Kessler et al., 2005; Moffitt et al., 2010; Somers et al.,
2006; Wittchen & Jacobi, 2005). The point prevalence of SAD has been estimated
to approximately 15% in the Swedish population (Furmark, et al., 1999; Furmark,
2002), and has internationally been estimated to be even higher among young
people (Heimberg, Stein, Hiripi & Kessler, 2000; Moffitt et al., 2010; Somers et
al., 2006). SAD is characterised by an intense fear or anxiety related to social
situations where the individual might be evaluated or be under scrutiny, that is out
of proportion to the actual threat and non-congruent with the sociocultural context
(APA, 2013a; Heimberg et al., 2014). Individuals suffering from SAD often fear
appearing or being judged as boring, unlikable, or that others will reject them, and
subsequently will avoid, or endure with great fear and anxiety, situations where
these fears might arise. Situations commonly described to evoke these fears are for
example: socialising, eating with others or giving a public performance. In the
fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (4th
ed., text rev.; DSM–IV–TR; American Psychiatric Association, 2000) one had to
specify if the social phobia (SP), as it was still known, was generalised or specific
and thus treated these as subtypes (APA, 2000). Generalised SP was characterised
by fearing “most” social situations, however due to debate regarding the
operationalisation of the term (Heimberg et al., 2014) the criteria were changed for
the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders
(DSM–5; American Psychiatric Association, 2013a) such that specifying SAD as
specific is only done if the anxiety is confined to performance situations, thus
removing the generalised subtype. SAD has a relatively early age of onset
(Hayward, Killen, Kraemer & Taylor, 1998; Stein & Gorman, 2001; Rosellini,
Page 9
2
Rutter, Bourgeois, Emmert-Aronson & Brown, 2013). Rosellini et al. (2013) found
that most individuals were diagnosed either before 10 years of age, 21%, or
between 14-17 years, 28,1%, whereas only 11,4 % were diagnosed after 23 years
of age. Because of the low rate of spontaneous remission SAD is said to follow a
chronic course if not treated (Beard, Moitra, Weisberg & Keller, 2010; Yonkers,
Dyck & Keller, 2001), and individuals with SAD reportedly have a lifetime
comorbidity with other mental disorders of between 62-92% (Fehm et al., 2005).
The most prevalent comorbid disorders are other anxiety disorders, substance
abuse and depression (Fehm et al., 2005), with the latter two appearing to develop
subsequently to SAD (de Graaf, Bijl, Spijker, Beekman & Vollebergh, 2003; Fehm
et al., 2005; Weiller, Bissorbe, Boyer, Lépine & van Lecrubier, 1996). Individuals
with SAD also tend to display lower workplace functioning, more unemployment
(Moitra, Beard, Weisberg & Keller, 2011), less quality of life, (Barrera & Norton,
2009; Olatunji, Cisler & Tolin, 2007; Sung et al., 2012; Wong, Sarver & Beidel,
2012), lower levels of educational and career attainment, and less relationship
satisfaction (Wittchen, Fuetsch, Sonntag, Müller & Liebowitz, 1999). Because of
this SAD is estimated to cost 136 million euro per million inhabitants (Acarturk et
al., 2009), albeit having the least estimated direct cost of the anxiety disorders, due
to the low levels of help seeking and low medical costs (Konnopka, Leichsenring,
Leibing & König, 2009). The societal cost of SAD is mostly related to indirect
factors such as higher unemployment and lower workplace functioning (Acarturk
et al., 2009; Konnopka et al., 2009). As a widespread disorder SAD could also be
considered an issue for the democratic process as people with SAD have a hard
time speaking up and asserting themselves, and an equality problem since more
women than men are diagnosed (Heimberg et al., 2000; Somers et al., 2006;
Wittchen & Jacobi, 2005) with a ratio of 3:2 (Furmark, 2002) and the differences
seem to be greater at younger ages (Gren-Landell et al., 2009; Wittchen et al.,
1999).
Cognitive Behaviour Therapy for Social Anxiety Disorder
As the current thesis is interested in the prediction of treatment response,
specifically to Internet-delivered Cognitive Behaviour Therapy (ICBT) for SAD, a
short review of the literature on Cognitive Behaviour Therapy and ICBT for SAD
is in order. The psychological treatment most widely studied in the treatment of
SAD, and anxiety disorders more widely, is CBT (Olatunji, Cisler & Deacon,
2010). In fact, Mayo-Wilson et al. (2014) performed a systematic review and
network meta-analysis and noted that there is some evidence for a differential
effect between CBT and other psychological treatments for SAD, such that CBT
appears to have a greater effect. Further, following a systematic literature review in
2005, the Swedish Council on Health Technology Assessment recommended CBT
Page 10
3
as the first choice of psychological treatment for SAD, and stated that there is a
strong scientific foundation on which to base the conclusion that CBT is an
effective treatment of SAD (Swedish Council on Health Technology Assessment
[SBU], 2005). CBT for SAD has been studied in group format, such as the
Cognitive Behavioral Group Therapy-model (CBGT; Heimberg & Becker, 2002),
and in individual format, such as the Individual Cognitive Therapy-model based on
the cognitive model of social anxiety proposed by Clark and Wells (1995). The
types of interventions used in CBT, regardless of format, in the treatment of SAD
are usually focused around exposure, behavioural experiments and cognitive
restructuring, interspersed with elements of psychoeducation and social skills
training (Rodebaugh, Holaway & Heimberg, 2004).
Since the late 1990s, researchers have been interested in the possibility of
delivering psychological treatment, particularly CBT, via the internet (Andersson,
Carlbring, Ljótsson & Hedman, 2013). The Swedish approach, most commonly
referred to as ICBT has been defined as “...a therapy that is based on self-help
books, guided by an identified therapist which gives feedback and answers to
questions, with a scheduling that mirrors face-to-face treatment, and which also
can include interactive online features such as queries to obtain passwords in order
to get access to treatment modules” (Andersson, et al., 2008, p. 164). The foremost
question has been whether or not ICBT produces treatment results similar to those
reported in studies of traditional face-to-face CBT. Hedman, Ljótsson and
Lindefors (2012) conducted a systematic review of published studies on CBT
delivered via the internet and reported that effect sizes were indeed comparable to
those reported in studies of traditional CBT for a multitude of psychiatric
disorders. Furthermore, in a systematic review and meta-analysis reviewing 13
studies comparing ICBT to similar face-to-face treatments, Andersson, Cuijpers,
Carlbring, Riper and Hedman (2014) found no difference in treatment effect when
aggregating results from studies focusing on SAD, panic disorder, depression,
body dissatisfaction, male sexual dysfunction and spider phobia.
Social anxiety disorder is one of the most frequently studied disorders in the ICBT
literature (Boettcher, Carlbring, Renneberg & Berger, 2013). Boettcher et al. in
their review, reported 21 studies on internet-based treatments of SAD from 4
different countries. Additionally, the authors report that of the 17 studies
investigating the efficacy of ICBT for SAD, 15 report large within-group effect
sizes (d > 0.80). A further example is provided by Arnberg, Linton, Hultcrantz,
Heintz and Jonsson (2014) who in their systematic review of the literature on
internet-delivered psychological treatments found 16 trials investigating treatment
for SAD, and rated the evidence of efficacy as being of moderate quality. The first
Page 11
4
Swedish study on ICBT for SAD was published in 2006 by Andersson et al., The
authors noted that there appears to be a reluctance to seek treatment among the
target group, due to the fear of embarrassment often associated with social anxiety,
and that ICBT therefore provides an opportunity to reach a group of patients who
previously would not have sought treatment. Andersson and colleagues reported a
large mean within-group effect size of d = 0.87. A follow-up of the original
participants five years later revealed that treatment gains at 1-year follow-up were
sustained at 5-year follow up (Hedman, Furmark et al., 2011). The 9-week
program used in the study by Andersson et al. (2006) has since been altered several
times, retaining the core content of the 9 modules (presented below under
Treatment) but varying for example length of the treatment and whether or not the
treatment is guided by a therapist (Furmark et al., 2009). A 15-week version of the
treatment was compared to 15 weeks of CBGT in a randomised controlled trial
conducted by Hedman, Andersson et al. (2011). The results showed that ICBT was
noninferior to the established CBGT-treatment. Furthermore the effectiveness of
the treatment has been investigated within the context of a regular care internet
psychiatry clinic (El Alaoui, Hedman et al., 2015). El Alaoui, Hedman et al.
studied a sample of 654 patients seeking psychiatric care for SAD during a 4-year
period, and who received ICBT from the specialised clinic. The authors reported a
large pretreatment to 6-month follow-up within-group effect of Cohen’s d = 1.15,
concluding that ICBT is an effective way to deliver treatment to patients with SAD
within the mental health care system.
Neuroimaging
Mapping the brain, and corresponding functions began in the middle of the 19th
century (Broca, 1865), a hundred years prior to the events leading up to functional
magnetic resonance imaging. Functional Magnetic Resonance Imaging (fMRI) is a
neuroimaging method that has, during the last decades, revolutionised the field of
neuroscience with the ability to noninvasively measure neuronal activity (Heeger
& Ress, 2002; Logothetis, Pauls, Augath, Trinath & Oeltermann, 2001) by way of
the blood oxygen level dependent (BOLD) signal (Amaro & Barker, 2006). The
first study using the BOLD signal to measure brain activity was reported in 1990
(Ogawa, Lee, Kay & Tank, 1990), and the use of BOLD fMRI rapidly grew to
become widespread and frequently used (Wager, Lindquist & Kaplan, 2007). It has
been used to study the brain function regarding a wide range of disciplines, from
the nature of consciousness (Lloyd, 2002), meditation (Cahn & Polich, 2006) to
evaluating the differential brain processes involved in mental disorders (Brooks &
Stein, 2015). fMRI uses the hemodynamic response, resulting from increased
regional metabolism, and its functional coupling to neuronal activity to
noninvasively measure brain activity (Heeger & Ress, 2002; Hipp & Siegel, 2015;
Page 12
5
Logothetis et al., 2001). It should also be noted that even if some aspects of the
relationship between the hemodynamic response and increased neuronal activity
have been adequately clarified, such that for example local increase of BOLD
signal is taken to mean greater neural activity (Heeger & Rees, 2002; Logothetis et
al., 2001), there are still some aspects that have not (Heeger & Rees, 2002; Hipp &
Siegel, 2015). The complexity of the relationship between the hemodynamic
response function and the BOLD signal requires equally complex methodology in
order to interpret the underlying neural activity (Heeger & Rees, 2002). In a
singular study by Hipp & Siegel (2015), the authors were able to show that the
correlation between BOLD response and neuronal activity varied depending on
several factors such as the individual subject, the brain area was being studied and
the frequency of the magnetoencephalography (MEG), providing firsthand
evidence of the functional coupling between neuronal activity and the
hemodynamic response as well as what might interfere with the approximation.
Neurobiology of Social Anxiety Disorder
The neurobiological study of SAD began in the late 1990s (Bell, Malizia & Nutt,
1999) and since then the number of studies in the field has grown rapidly and these
in turn have begun to elucidate the neurobiological underpinnings of the disorder
(Brühl, Delsignore, Komossa & Weidt, 2014; Etkin & Wager, 2007; Freitas-Ferrari
et al., 2010; Furmark et al., 2002). As the current thesis aims to evaluate the
predictive value of said neurobiological underpinnings, these will be briefly
reviewed below. As previously mentioned SAD is characterised by an intense fear
of social situations. In the study of fear acquisition and its expression the amygdala
(LeDoux, 2003), the insula, the hippocampus and the anterior cingulate cortex
(ACC) have been found to be of critical importance (Fullana et al., 2016) with
further indications of the involvement of the prefrontal cortex (PFC) and especially
the medial prefrontal cortex (mPFC) and the dorsal anterior cingulate cortex
(dACC) (Etkin, Egner & Kalisch, 2011; Milad & Quirk, 2012). Generally an
increase in functional reactivity in the above mentioned regions has been reported
in anxiety disorders, forming a fear circuitry (Etkin & Wager, 2007; Etkin et al.,
2011; Shin & Liberzon, 2010), however the results are not unanimous. In a recent
meta-analysis by Brühl et al. (2014), increased functional reactivity in SAD
compared to healthy controls was found in the amygdala, the insula, the
hypothalamus, the ACC, the occipitotemporal regions, the medial and ventrolateral
prefrontal cortex (mPFC, vlPFC) and the parietal cortex, with less consistent
findings related to increased activation in the fusiform gyrus and the hippocampus,
and ambiguous findings regarding the dorsolateral prefrontal cortex (dlPFC).
Furthermore, increased connectivity between the frontal areas and the thalamus,
the amygdala and the occipital cortex (OCC) was found, as well as decreased
Page 13
6
connectivity between the frontal areas of the two hemispheres (Brühl et al., 2014).
These findings are in line with and expand upon previous reviews (Freitas-Ferrari
et al., 2010). The areas previously implicated as being most connected to the fear
circuitry and SAD have been the amygdala, the insula, the ACC and the PFC why
these will be reviewed individually.
The Amygdalae (Latin for almond) are two almond shaped cell masses located
subcortically in the human medial temporal cortex (Davis & Whalen, 2001). The
amygdala is considered central to the human emotional system, being implicated in
emotional learning, memory, inhibition and regulation as well as influencing
attention, perception and social responding (Phelps & Ledoux, 2005). The
expansive interest in the amygdala originated from its involvement in fear
acquisition and expression, known as fear conditioning, a form of classical or
pavlovian conditioning (Phelps & Ledoux, 2005). In fear conditioning, a neutral
stimulus (NS) comes to elicit a defensive behaviour with it’s adhering
physiological responses after being associated with an aversive event (US), thus
becoming a conditioned stimulus (CS; Rescorla & Wagner, 1972). Amygdalar
reactivity has long been implicated in SAD (Bell, Malizia & Nutt, 1999; Brühl et
al., 2014), for example in response to the processing of facial expressions depicting
neutrality (Cooney, Atlas, Joormann, Eugene & Gotlib, 2006), anger (Evans et al.,
2008), fear (Blair, Shaywitz et al., 2008; Blair, Geraci, Korelitz et al., 2011; Phan
et al., 2013; Prater, Hosanagar, Klumpp, Angstadt & Phan, 2013), as well as
generally having an increased response related to face processing and perception
(Gentili et al., 2008). Furthermore increased amygdalar activity, relative to healthy
controls, has been observed for individuals with SAD in response to negative
comments (Blair, Geraci et al., 2008), social anxiety-related words (Schmidt,
Mohr, Miltner & Straube, 2010), verbal stories regarding social transgressions
(Blair et al., 2010) viewing negative non-social images (Shah, Klumpp, Angstadt,
Nathan & Phan, 2009) and when experiencing anticipatory anxiety (Boehme et al.,
2013; Brühl et al., 2011).
The insular cortex, in humans, is a hidden lobe situated in the depths of the Sylvian
fissure where it takes up less than 2% of the total cortical surface area
(Nieuwenhuys, 2011). The insula receives input from certain sensory thalamic
nuclei, and is reciprocally connected to the amygdala (Kohn et al., 2014), as well
as many other limbic and cortical areas, and is involved in speech production,
processing of emotions, and pain (Nieuwenhuys, 2011). Thus the insular cortex is
critical for interoceptive awareness (Ronchi et al., 2015) and self-consciousness
(Craig, 2011; Smith & Lane, 2015), and an excessive attention to these processes is
a principal characteristic of SAD (APA, 2013b; Clark & Wells, 1995). In healthy
Page 14
7
controls higher activation of the insular cortex has been linked to anxiety (Carlson,
Greenberg, Rubin & Mujica-Parodi, 2011; Stein, Simmons, Feinstein & Paulus,
2007). In SAD the insula and amygdala are generally simultaneously activated
(Boehme et al., 2013; Klumpp, Angstadt, Nathan & Phan, 2010; Gentili et al.,
2008; Kohn et al., 2014; Schmidt et al., 2010; Shah et al., 2009), and thus the
insula shows increased reactivity related to facial expressions (Klumpp et al., 2010;
Gentili et al., 2008), negative imagery (Shah et al., 2009), anticipatory anxiety
(Boehme et al., 2013) and social-anxiety related words (Schmidt et al., 2010). A
normalisation of blood flow in the insula has been shown in individuals who have
undergone CBT treatment for SAD (Klumpp, Fitzgerald & Phan, 2013), and when
using emotional regulation strategies derived from CBT (Brühl, Herwig,
Delsignore, Jäncke & Rufer, 2013).
The ACC is a large region surrounding the frontal part of the corpus callosum, in
the medial walls of the frontal lobule (Devinsky, Morell & Vogt, 1995; Etkin, et
al., 2011). Further the ACC has been implicated in the commitment to a course of
action and the evaluation of costs and alternative options (Kolling, Behrens,
Wittmann & Rushworth, 2016). More precisely, studies have found the ACC to be
crucial in the recruitment of cognitive control (Kerns et al., 2004), reinforcement
learning (Kennerley, Walton, Behrens, Buckley & Rushworth, 2006) and
emotional processing, including both appraisal and expression of emotion (Etkin et
al., 2011). The ACC is also extensively connected to the amygdala (Etkin et al.,
2011) and there are indications that these connections may be aberrant in
individuals with SAD, with decreased connectivity to dACC (Prater et al., 2013)
and increased connectivity to the rostral ACC (Demenescu et al., 2013).
Interestingly however, lower dACC-amygdala coupling at pretreatment has also
been related to greater improvement after treatment (Månsson et al., 2015).
Further, increased activity in the ACC has been implicated in SAD in response to
the processing of facial expressions (Blair, Shaywitz et al., 2008; Blair, Geraci,
Korelitz et al., 2011; Klumpp, Post, Angstadt, Fitzgerald & Phan, 2013;
Labuschagne et al., 2012), processing of aversive images (Gaebler, Daniels,
Lamke, Fydrich & Walter, 2014) as well as processing of self-referential
comments (Blair, Geraci, Otero et al., 2011). However decreased activity has been
observed in the dACC as compared to healthy controls in relation to viewing angry
facial expressions (Evans et al., 2008).
The PFC is a large region in the frontalmost part of the neocortex, underlying
higher order cognition and executive skills, in addition to being extensively
connected to other areas such as the ACC, the amygdala, and the parietal- and
occipital areas (Wood & Grafman, 2003). The mPFC has been shown to play a
Page 15
8
major role in inferring the best strategy to maximise reward in a given setting
(Donoso, Collins & Koechlin, 2014), and has also been extensively evaluated in
the extinction of learned fear response (Lissek et al., 2014) and regulation of
activity in the amygdala amongst other areas (Motzkin, Philippi, Wolf, Baskaya &
Koenigs, 2015). An increased activity has also been shown in the mPFC when
participants with SAD process self-referential comments from a second person
viewpoint (Blair, Geraci, Otero et al., 2011), when they observe fearful facial
expressions (Frick, Howner, Fischer, Kristiansson & Furmark, 2013; Klumpp,
Angstadt & Phan, 2012), observe sad facial expressions (Labuschagne et al.,
2012), when they process social-anxiety related words (Schmidt et al., 2010), hear
angry voices (Quadflieg, Mohr, Mentzel, Miltner & Straube, 2008), view harsh
faces (Ziv, Goldin, Jazaieri, Hahn & Gross, 2013) and read verbal stories regarding
social transgressions (Blair, Geraci, Otero et al., 2011). Furthermore, decreased
activity has been shown when processing angry facial expressions (Phan et al.,
2013), processing self-referential comments in first person viewpoint (Blair,
Geraci, Otero et al., 2011) as well as when playing a trust game (Sripada et al.,
2009). The dlPFC has been implicated in the evaluation of emotions (Etkin, 2010)
and mental states (Ochsner, Silvers & Buhle, 2012) and has been proposed to gate
the conscious awareness of emotions (Etkin, 2010). In SAD an increased reactivity
has been observed in the dlPFC compared to controls when observing fearful facial
expressions (Blair, Shaywitz et al., 2008), angry facial expressions (Evans et al.,
2008), viewing facial expressions with higher intensity (Klumpp, Angstadt, Nathan
& Phan, 2010), but also a decreased activation when viewing facial expressions in
general (Gentili et al., 2008) Furthermore higher activity in the dlPFC has been
related to increased anxiety symptoms (Koric et al., 2012; Pujol et al., 2013) with
negative expectation (Brühl et al., 2011) responding to verbal criticism (Blair,
Geraci et al., 2008) as well as during a trust game (Sripada et al., 2009).
Successful treatment of SAD has been shown to impact reactivity in several areas,
with CBT attenuating activation in the insula, the vmPFC (Klumpp, Fitzgerald &
Phan, 2013), amygdala (Furmark et al., 2002; Goldin & Gross, 2010; Månsson et
al., 2013) and the posterior superior temporal gyrus (Goldin et al., 2014).
Additionally CBT has been shown to increase activation in the dlPFC, the dmPFC
(Goldin, Ziv, Jazaieri, Hahn, Heimberg & Gross, 2013), and the mPFC (Goldin et
al., 2014) as well as alter the frontal-amygdalar connectivity (Goldin et al., 2013;
Månsson et al., 2013). Thus psychological interventions seem to have a strong
impact on the neural reactivity and connectivity related to SAD, however as of yet
there is little consensus regarding which areas have the most impact on treatment
change and related neural activity.
Page 16
9
Self-related processing and neural-correlates
In this thesis we aim to investigate a process proposed to be central to SAD,
namely the fear of negative evaluation, by applying a paradigm which presents
either self- or other-referential criticism to the subject. Generally studies have
shown neural activation in most brain areas during self-related processing
(Northoff et al., 2006) though it seems that emotional processing related to the self
may elicit greater activation in the central midline structures such as the vmPFC,
the ACC, the dmPFC, and the PCC (van der Meer, Costafreda, Aleman & David,
2010) but also areas including the temporoparietal junction, the precuneus, and the
temporal poles (Legrand & Ruby, 2009).
Amygdalar and insular activity has further been implicated as having a central role
in self-related emotional processing, as opposed to appraisal of the self (Northoff et
al., 2006). In spite of the above mentioned findings, heavy critique has been
directed at the field of imaging research studying self-related processing (Legrand
& Ruby, 2009) due to conclusions being drawn based on neural correlates.
Legrand and Ruby instead proposed that self-evaluative processing does not draw
upon a unique network per se, but relates any object to the subject, thus illustrating
that the areas are involved in inferential processing and memory recall. In line with
the above, the same network has been implicated in processing of personally
familiar stimuli, general evaluation (Northoff, Qin & Feinberg, 2011) as well as
sharing activation patterns with parts of the Default Mode Network (DMN; Qin &
Northoff, 2011). Furthermore because of the relative equivalence between the
DMN and areas related to self-referential processing it has been argued that the
“self” might be largely indistinguishable from the DMN (Northoff, Qin &
Feinberg, 2011).
In a study by Blair, Geraci et al. (2008) 32 comments of emotional- negative,
neutral or positive valence were directed either in second-, you, or third person, he
or she, to individuals with SAD and healthy controls. Significant interactions for
group, valence and referent were found in the mPFC and the bilateral amygdala,
thus individuals with SAD showed significantly greater BOLD response to
negative comments that were self-referential, indicating hyperactivity in SAD
regarding the circuitry underlying the neural representation of the self. Similar
paradigms have been used since, building upon the original used in Blair, Geraci et
al. (2008), amongst others Blair, Geraci, Otero et al. (2011), Abraham et al., (2013)
and Månsson et al. (2016). Boehme, Miltner, and Straube, (2015) also found, in
line with the above, that individuals, with higher social anxiety, had a positive
correlation between more self-focused attention and activations of the mPFC, the
PCC, the temporoparietal junction and the right insula in a self-focusing paradigm.
Page 17
10
Prediction of treatment response
Almost a quarter of a century ago, Steketee and Chambless (1992) noted that
clinicians, rather than boasting the success of psychological treatments, should
look to explore the reasons why treatments fail to attain the expected results. This
is particularly true of treatments such as CBT and ICBT for SAD, on which plenty
of research has been done, yet approximately half of RCT-participants are
classified as non-responders (Loerinc et al., 2015). Prediction research therefore
aims to ascertain which patient characteristics are related to treatment outcome, so
as to facilitate treatment development and selection. Besides treatment outcome,
defined as the posttreatment or follow-up level of functioning of the individual,
measured for example by whether or not the individual still meets diagnostic
criteria for the disorder in question, prediction research is also concerned with
predicting treatment response, defined as whether or not the individual has
responded adequately to the treatment given (Steketee & Chambless, 1992). A
great many variables have been studied as potential predictors of treatment
outcome or treatment response in CBT for SAD, but few results have proven to be
robust. Eskildsen, Hougaard and Rosenberg (2010) compiled a review of 28
studies investigating patient characteristics at pretreatment and their predictive
value in regard to treatment response. One of the main conclusions from the review
was that “...there is little clinically or theoretically relevant knowledge to be gained
from existing studies of pretreatment patient variables as predictors of dropout and
treatment outcome in CBT for patients with [SAD]” (Eskildsen, Hougaard &
Rosenberg, 2010, p. 103). Concerning demographic characteristics, Eskildsen,
Hougaard and Rosenberg report only one study investigating their relationship to
treatment outcome, in which being female and having a higher education was
related to better treatment outcome. The predictive value of age has also been
investigated, and in a couple of studies where participants treated with ICBT
higher age has been related to better treatment outcome (Kawaguchi et al, 2013)
and faster rate of change (El Alaoui, Ljótsson, et al., 2015). Further review will
focus on the two clinical characteristics most studied as regards prediction of
treatment outcome in SAD, namely comorbid depression and pretreatment
symptom severity of SAD. The predictive value of treatment expectancy will also
be reviewed, due to its relative consistency as a predictor of treatment response.
Depression
Depression has long been studied as a possible risk factor for poor treatment
response and poor treatment outcome when delivering CBT for SAD. Chambless,
Tran & Glass (1997) found that higher levels of self-rated depression were related
to lower levels of treatment response at 6 months after treatment. These results
were partially replicated a couple of years later by Scholing and Emmelkamp
Page 18
11
(1999), who reported that severity of depressive symptoms at pretreatment were
related to higher levels of social avoidance at posttreatment. The relationship was
however not statistically significant at 18 month follow up. Eskildsen, Hougaard
and Rosenberg (2010) in their systematic review reported that there appears to be
some support for the findings that depressive symptoms at pretreatment are
negatively related to treatment outcome. The authors do however express that the
relationship between depressive symptoms at pretreatment and degree of change is
still in need of further clarification, as two of the studies reviewed in fact reported
a positive relationship. In a more recent study, Fracalanza, McCabe, Taylor and
Antony (2014) studied the effects of mood disorder and depressive symptom
severity on rate of change when receiving face-to-face CBT for SAD. The authors
reported no significant relationship between depressive symptoms and rate of
change. Depression was however related to severity of SAD-symptoms at pre- as
well as posttreatment which Fracalanza et al. interpret as being in line with
previous findings.
The effects of depression on treatment response and treatment outcome have also
been studied in regard to ICBT, and the findings appear similar to those from
research on face-to-face CBT. Hedman et al. (2012), studying outcome
determinants of CBGT and ICBT respectively, reported that depression, as well as
general anxiety, significantly moderated the relationship between treatment and
outcome, such that lower levels of depression were related to better outcome in
ICBT, but not in CBGT. The authors interpret the findings as possible evidence of
ICBT perhaps being better suited for patients with lower degrees of general anxiety
and depression, as these patients have a greater ability to independently make plans
and follow through on said plans. In contrast to the above findings El Alaoui,
Ljótsson et al. (2015), studying a sample of 726 outpatients with SAD receiving
ICBT, found no statistically significant relationship between depressive symptoms
and treatment response. One possible explanation for the ambiguous findings
regarding depression and response to CBT-treatment is provided by Tillfors,
Furmark, Carlbring and Andersson (2015). The authors studied potential risk
profiles for diminished treatment response to ICBT for SAD. Tillfors and
colleagues applied a cluster analytic strategy to an aggregated dataset containing
data from five RCT studies (ntotal = 167) to identify subgroups of social avoidance
and depression that were related to an increased risk of poor treatment response.
One high risk cluster was identified, containing patients with high levels of
avoidance and high levels of depression. The authors therefore conclude that high
levels of depression and high levels of social avoidance by themselves do not
constitute risk factors of poor treatment response, but participants in which the two
are combined tend to run a higher risk of poor treatment response.
Page 19
12
Pretreatment symptom severity of Social Anxiety Disorder
Generally in research on prediction of psychotherapy outcome, pretreatment
symptom severity has been found to be related to treatment outcome, such that
higher levels of pretreatment symptoms predict worse treatment outcome
(Lindhiem, Kolko & Cheng, 2012), and this appears also to be the case regarding
CBT for SAD. As an example, Kawaguchi et al. (2013) studied predictors of
treatment outcome in a naturalistic sample of outpatients receiving CBGT for SAD
and found a statistically significant relationship between pretreatment SAD and
posttreatment status. In the previously mentioned study by Hedman et al. (2012),
the same relationship between pretreatment SAD severity and outcome was
observed in both CBGT and ICBT, demonstrating that the relationship appears to
hold true regardless of treatment. There are however studies in which contradicting
results have been found. In the studies by Chambless, Tran and Glass (1997) as
well as Schibbye et al. (2014), pretreatment SAD severity was found to be
unrelated to treatment outcome. Another interesting relationship is the one between
pretreatment SAD severity and rate of change, or treatment response. Nordgreen et
al. (2012), in line with the aforementioned findings, reported that ratings of
symptom severity at pretreatment were negatively related to diagnosis-free status
at follow-up 6 to 12 months later, but positively related to reliable change
measured with the Reliable Change Index (RCI; Jacobson & Truax, 1991).
Interestingly, in the naturalistic study by El Alaoui, Ljótsson et al. (2015), a large
and methodologically well suited study as pertains assessing the relation between
pretreatment symptom severity and treatment response, no statistically significant
relationship was found between pretreatment SAD severity and either treatment
outcome or treatment response, suggesting that pretreatment severity of symptoms
has no effect on treatment response or outcome in outpatients receiving ICBT.
Treatment expectancy
One of the more consistently reported predictors of treatment response and
treatment outcome concerning CBT as well as ICBT for SAD, is treatment
expectancy (Eskildsen, Hougaard & Rosenberg, 2010; Boettcher, Renneberg &
Berger, 2013). Boettcher, Renneberg and Berger (2013) define treatment
expectancy as “... the individual change a client expects in the course of therapy”
(p. 204), and the related concept of treatment credibility as “... how logical,
beneficial and trustworthy a treatment seems” (p. 204). Price and Anderson (2012)
studied the predictive value of treatment expectancy in the treatment of public
speaking fears. The authors compared two treatments, CBGT and individual face-
to-face CBT containing elements of virtual reality exposure, to ascertain the
relationship between treatment expectancy and rate of change. The reported results
provided support for a medium-sized to large positive relationship between
Page 20
13
treatment expectancy and rate of change irrespective of treatment. Another
interesting example is provided in the study by Borge, Hoffart and Sexton (2010)
examining predictors of treatment outcome in residential CBT and residential
interpersonal psychotherapy. The authors reported treatment expectancy, as well as
age of onset, as being the most reliable predictors of treatment outcome in both
treatments.
Concerning ICBT, Boettcher, Renneberg and Berger (2013) explored the
relationship between treatment expectancy, measured at week two of treatment,
and symptom reduction following ICBT in a sample of 109 participants with SAD.
The authors reported a statistically significant relationship such that higher
treatment expectancy was related to greater symptom reduction. Furthermore the
relationship between positive treatment expectancy and greater symptom reduction
was mediated by early gains (symptom reduction from week 0 to week 2).
Boettcher, Renneberg and Berger also report a statistically significant positive
relationship between treatment expectancy and treatment adherence. A limitation
noted by the authors of the aforementioned study is however the high and atypical
dropout rate of 37%. Their results are nevertheless corroborated by the previously
mentioned studies by Hedman et al. (2012) and El Alaoui, Ljótsson et al. (2015),
the first showing a positive relationship between treatment expectancy and
treatment response in a research sample of 126 participants receiving ICBT, and
the second showing a similar effect in a sample of 726 patients receiving ICBT
through outpatient care.
Biological predictors of treatment response
Barber (2007) noted that biological variables such as genetic polymorphisms and
brain imaging data had seldom been investigated as potential predictors of
treatment outcome in psychotherapy research. Two studies have investigated the
genetic outcome determinants of CBT for SAD namely Andersson et al. (2013)
and Hedman et al. (2012), who both report no statistically significant relationship
between the examined allelic variations and long-term treatment outcome. The
literature on brain imaging data and prediction has however produced more
encouraging results. The first report suggesting attenuated amygdalar reactivity as
a predictor of treatment response was reported by Furmark et al., (2002) who used
positron emission tomography to show that treatment response to both CBT and
pharmacological treatment for SAD was related to attenuated regional cerebral
blood flow (rCBF) in the amygdala during an anticipatory anxiety task.
Furthermore the authors reported that attenuated rCBF in the amygdala, the
periaqueductal gray area, and the left thalamus was predictive of responder status
at one year follow up, implicating a major role for activity related to these areas in
Page 21
14
the continuation of the disorder. fMRI-data have also been studied to ascertain if
activation patterns related to certain paradigms can be used to predict treatment
response in SAD (Doehrmann et al., 2013; Klumpp, Fitzgerald & Phan, 2013) as
well as other anxiety disorders (e.g. Ball, Stein, Ramsawh, Campbell-Sills &
Paulus, 2014) and depression (e.g. Siegle, Carter & Thase, 2006). In regard to
SAD, Doehrmann et al. (2013) reported that activation in the ventral and dorsal
occipitotemporal cortex as participants were exposed to images depicting angry
faces predicted treatment response, such that greater activation predicted greater
treatment response. Using the combination of imaging data and pretreatment
symptom severity the authors were able to account for 40% of the variation in
treatment response. Furthermore, Klumpp, Fitzgerald and Phan (2013) were also
able to use fMRI-data to predict treatment response to CBT for SAD. This time the
participants were exposed to images depicting faces showing fear, anger, happiness
or neutrality, and the authors reported that activation of brain regions involved in
visual processing when viewing angry and fearful faces, and activation in the
dACC and the dmPFC when viewing fearful faces, predicted treatment response.
Additionally, Klumpp, Keutmann, Fitzgerald, Shankman and Phan (2014) used
resting state fMRI images from the same sample to assess the predictive value of
functional connectivity. The authors reported that connectivity between the
amygdalae and the prefrontal regions at pretreatment was significantly correlated
with treatment response. In addition to standard univariate methods, methods of
multivariate pattern analysis (MVPA) have also been used to examine if
pretreatment brain imaging data can predict treatment response. Using said
approach to predict treatment response to CBT for SAD, Månsson et al. (2015)
used MVPA in an attempt to correctly classify participants as either responders or
non-responders depending on BOLD activation to a previously mentioned fMRI-
paradigm consisting of self- or other referential criticism. The authors reported a
91.7% rate of successful classification, with a specificity of 100% and a sensitivity
of 83.3%, based on activation in the dACC and the amygdala. Whitfield-Gabrieli et
al. (2015) used MVPA on the same sample as the previously described study by
Doehrmann et al. (2013) to assess if patterns of connectomic measures could
predict treatment response to group-based CBT for SAD. Whitfield-Gabrieli et al.
(2015) reported that connectomic measures together with a measure of
pretreatment symptom severity accounted for 60% of the total variance in
treatment response, and that using these data they were able to correctly predict
whether or not the individual would be classified as a treatment responder in 81%
of cases, with a specificity of 78% and a sensitivity of 84%.
Page 22
15
Aims
Despite a number of published studies investigating the relationship between a
variety of predictor variables and treatment response, as well as treatment outcome,
in CBT for SAD, few consistent findings have been reported. The current thesis
therefore aims to explore predictors of change during treatment in a sample of
participants receiving ICBT for SAD. In particular we wish to explore the
predictive properties of fMRI-data collected 9 weeks prior to treatment onset,
using a disorder-relevant paradigm which has previously been used to predict
treatment response. The paradigm, a modified version of the paradigm used in
Blair, Geraci et al. (2008), consists of presenting negative social information to the
participants, and represents a proposed fundamental part of SAD, namely the fear
of negative evaluation. Further elucidating the neural correlates of these fears is of
critical importance to the future understanding of SAD. Using neuroimaging data
to predict treatment response provides a possibly more objective and reliable way
of predicting who will benefit from treatment, and so facilitating effective resource
management in the hard pressed context of psychiatric care. Thus our hypotheses
were as follows, stated as the alternative hypotheses: (1) whole-brain analyses will
reveal greater activation in areas related to the fear circuitry when viewing self-
referential criticism, rather than other-referential criticism. (2) activity in areas
related to the fear circuitry which have previously shown aberrant activity in SAD,
namely the amygdala, the insula, and the ACC, will be significantly related to
treatment response (3) the combination of brain measures and behavioural
measures will explain significantly more of the variation in treatment response than
behavioural measures alone.
Materials and methods
General procedure
The current thesis was part of a larger research project. The purpose of the research
project is to elucidate the mechanism behind CBT for SAD and thus how the brain
is affected and mediates treatment success. During the study, four MRI-scans were
conducted for each participants, at first baseline, second baseline which
corresponded to treatment onset, mid-treatment and posttreatment. Between the
first and second baseline there was a wait period of 9 weeks so as to enable a
within-group wait list control for the research project. The post treatment MRI-
scan was conducted soon after treatment completion, and posttreatment severity
measurements used in the current thesis were collected prior to the final MRI.
Procedure
The study was approved by the local ethics committee. Screening and inclusion
procedures were similar to those from a previous study by the same research group
Page 23
16
(Månsson et al., 2013). The participants were recruited via media advertisement. A
total of 265 individuals reported interest in the study on the webpage and 226
individuals answered all self-report questionnaires; demographics, MRI-safety
criteria, Liebowitz Social Anxiety Scale Self-Report (LSAS-SR; Baker, Heinrichs,
Kim & Hofmann, 2002), and Montgomery-Åsberg Depression Rating Scale
(MADRS-S; Montgomery & Asberg, 1979) see Figure 1. The participants had to
be at least 18 years of age, have no neurological disorder, have no metal implants
or health issues making the MRI hazardous, not be enrolled in any concurrent
psychological treatment, and if undergoing pharmacotherapy, having had a stable
dose for at least three months as well as agreeing to maintain a stable dose across
the duration of the study. A total of 113 participants were eligible after the initial
Online registration for eligibility
(n=265)
Did not complete all surveys (n=39)
Baseline-1 (n=51)
MRI
LSAS-SR
MADRS-S
Baseline-2 (n=46)
CEQ
Post treatment (n=46)
LSAS-SR
Excluded (n=62)
Comorbidity (n=14)
Ineligible for MRI (n=13)
Practical reasons (n=11)
Not SAD, or mild symptoms (n=9)
Withdrawal (n=4)
Suicidality (n=3)
Other treatment (n=3)
Menopause (n=3)
Neuropsychiatric disorder (n=2)
Complete online registration (n=226)
Telephone interview (n=113)
Excluded (n=5)
Pathological findings (n=3)
Withdrawal (n=2)
Figure 1
Flowchart regarding the recruitment and procedure of the study. MRI, Magnetic Resonance
Imaging; LSAS-SR, Liebowitz Social Anxiety Scales – Self Report; MADRS-S, Montgomery-
Åsberg Depression Rating Scale; CEQ, Credibility Expectancy Questionnaire
Page 24
17
screening and were subsequently interviewed by either master level psychologist
students or psychologists via telephone using the full Mini-International
Neuropsychiatric Interview (M.I.N.I.) version 7.0, developed by Sheehan et al.
(1998), and the social phobia section of the Structural Clinical Interview for DSM-
4 - Axis I Disorders (SCID-I; First, Gibbon, Spitzer & Williams, 1997).
Furthermore the participants were asked about any previous and contemporary
treatments, whether or not they had entered menopause, their reason for applying,
whether they met the safety criteria for MRI, and whether they experience
claustrophobia. Participants were required to meet diagnostic criteria for SAD
according to the Diagnostic and Statistical Manual of Mental Disorders, 5th edition
(DSM-5; APA, 2013a), have SAD as their primary diagnosis, have no current
severe depression, as indicated by a score above 20 points on MADRS-S, or
suicidal tendencies, as indicated by a score above 2 points on MADRS-S item 9,
not show claustrophobic tendencies, not meet the DSM-5 criteria (APA, 2013a) as
indicated by M.I.N.I. for either: current major depressive disorder, suicidality, any
form of bipolar syndrome, any psychotic syndrome, alcohol disorder or substance
use disorder, any eating disorders, or antisocial personality disorder. Prior to the
first MRI scan the participants were asked to sign forms of informed consent for
genetic analysis, the use of personal information according to the privacy
protection law and that the individual had read the information on the website
about the study procedure. Posttreatment assessment of symptom severity was
completed by participants upon treatment completion, but prior to attending the
fourth and final imaging session.
Participants
Of the initial 265 individuals who screened online for eligibility, a total of 50 were
thereafter included to be analysed. However due to an incidental finding, meaning
in the current context an unexpected abnormality in the individual's brain, during
the first baseline measurement another was recruited, thus 51 participants
underwent MRI-scanning at baseline 1. In total three participants were excluded
due to incidental findings. Furthermore another two participants withdrew before
commencing treatment, and have also been excluded from data analysis, and apart
from these no data-points used for the current analyses are missing. Demographics
characteristics of the current sample are presented in Table 1.
Page 25
18
Table 1:
Demographic description of the participants
Variable Cohort
Gender
Women
Men
29 (63%)
17 (37%)
Age
M (SD)
Min-max
30.7 (8.3)
19-52
Handedness
Right
Left
40 (87%)
6 (13%)
Education
Completed primary education
Completed secondary education
Completed vocational education
Ongoing university studies
Completed university studies
3 (7%)
7 (15%)
2 (4%)
16 (35%)
18 (39%)
Employment
Full time studies
Full time employment
Part time employment
No employment
21 (46%)
18 (39%)
5 (11%)
2 (4%)
Civil State
Married/Cohabiting with children
Married/Cohabiting without children
Non-cohabiting relationship without children
Single with children
Single without children
Other
16 (35%)
10 (22%)
4 (9%)
4 (9%)
9 (20%)
3 (7%)
Note. N = 46
Behavioural measures
LSAS-SR is an instrument developed to measure symptoms of social anxiety, it
consists of 24 items listing different social situations, where the individual is asked
to rate fear and avoidance of said situation on a 4-point likert-type scale (Baker et
al., 2002). The 12-week test-retest reliability is r = 0.83 (p < 0.01) for the total
score (Baker et al., 2002; see Appendix A for full questionnaire). MADRS-S is an
Page 26
19
instrument developed to measure symptoms of depression, consisting of 9 items
concerning mood, anxiety, sleep, appetite, concentration, initiative, emotional
engagement, pessimism and suicidality, each item is rated on a 7 point likert-type
scale, the one-week test-retest reliability has been reported to be cronbach’s 𝛼=
0.84 (Fantino & Moore, 2009), and the instrument has also been found to have a
high internal-reliability cronbach’s 𝛼 > 0.8 (Cunningham, Wernroth, Von
Knorring, Berglund & Ekselius, 2011; see Appendix B for full questionnaire). The
Credibility/Expectancy Questionnaire (CEQ; Borkove & Nau, 1972), measures the
individual's expectancy of a positive outcome and the perceived credibility of a
treatment. Internal consistency has been reported to be high, standardised
𝛼 between 0.84 and 0.85, and test-retest scores of 0.82 for the expectancy scale,
and 0.75 for the credibility scale (Devilly & Borkovec, 2000). In the current study
five items on the CEQ were used, rated on a 10 point likert-type scale, omitting the
last question from Devilly, and Borkovec (2000; see Appendix C for full
questionnaire).
Treatment
Participants were treated via a nine week guided ICBT self-help program. The
treatment has repeatedly been shown to be efficacious in the treatment of SAD,
with large effect sizes, and sustained treatment effects at five years after treatment
having been reported (Andersson, Carlbring, Furmark, 2012; Andersson et al.,
2006; Carlbring et al., 2007; Furmark et al., 2009; Hedman, Furmark et al., 2011).
The treatment consisted of nine modules, a new module being introduced once
every week, containing information, worksheets and homework. The first module
presents an introduction to CBT and SAD, modules 2-4 describe Clark & Wells
cognitive model of SAD (Clark & Wells, 1995) and restructuring of dysfunctional
cognitions, modules 5-7 introduce and encourage exposure with response
prevention, and modules 8-9 aim to increase the participants’ social skills and
promote relapse prevention. The treatment also exists as a self-help book
(Furmark, Holmström, Sparthan, Carlbring & Andersson, 2013).
Experimental task
The experimental task which participants performed while in the fMRI-scanner
consisted of viewing sentences of self-referential and other-referential criticism,
using a heavily modified version of a disorder-relevant paradigm initially
developed by Blair, Geraci et al. (2008). A similar version of this paradigm has
previously been used by the research group responsible for the current study to
predict treatment outcome (Månsson et al., 2015). The paradigm is an event-related
design and includes 27 criticising statements either self-referential, “Nobody likes
you”, or other-referential, “Nobody likes [significant other]”, where the other-
Page 27
20
referential target was a significant other which the subject had previously stated
that they were close to and who meant much to them (all statements used can be
read in Appendix D and translated to English in Appendix F). Participants were
instructed to read the sentences and press an arbitrary button with their right hand,
alternatively their left hand if they specifically requested it, once they had finished
reading the sentence, after which the sentence disappeared and a fixation cross
appeared as described below and in Figure 2. The instruction for the paradigm was
given in the waiting room prior to the fMRI scan. The sentences were randomised
in their order of presentation, and 36 sentences were displayed for each target, self
or significant other, these were randomly drawn from the respective 27 sentence
pool. The stimuli were presented for a maximum of 2500ms whereupon a fixation
cross with a duration of 500ms was presented prior to the next stimuli. The
sentences were interspersed with 45 fixation crosses with 30 having a duration of
2500ms and 15 a duration of 3500ms. Stimuli were presented using the E-prime
2.0 software (Psychology Software Tools, Pittsburgh, PA, USA), projected on a
screen and viewed through a tilted mirror attached to the head coil. The maximum
duration of the task was five minutes and 50 seconds, but depended on the
participants’ response time. If participants failed to respond to multiple sentences it
was possible for the paradigm to complete before all sentences had been
administered. See Figure 2 for illustration of the paradigm.
Data acquisition and preprocessing
Image acquisition was performed using a 3T General Electric MRI scanner with a
32 channel head coil (GE Medical Systems, Madison, Wisconsin, USA).
Functional T2-weighted BOLD images were acquired with an echo time of 30ms, a
repetition time of 2000ms, an 80° flip angle, a 250x250mm2 field of view
, an
Figure 2
Illustrating the negative statements. The fixation crosses with 500 ms duration was presented
prior to any stimuli. 36 sentences were displayed for each target, self or significant other, and
were randomly drawn from the respective 27 sentence pool. The random duration of the fixation
crosses was either 2500 ms or 3500 ms and there were 30 respectively 15 of each. “X” was
displaced by the name of the significant other for the participant.
Page 28
21
acquisition matrix size of 96x96 and an in-plane resolution of 2.6x2.6mm. Each
slice had a thickness of 3.4mm and 37 slices were acquired for each volume. For
the experimental paradigm described above, 175 volumes were collected per run,
and 10 dummy scans were run prior to image acquisition so as to avoid signal
detection due to progressive saturation. T1-weighted structural images were also
acquired. These images consisted of 180 slices with a voxel size of 0.5x0.5x1mm3
and a field of view of 250x250mm2 and were used for coregistration.
Quality control of the functional images was done manually by both authors via
ocular inspection of the mean image for each time series, as well as by analysing
time series difference data. This was done to identify any obvious artefacts or
quality reductions in the imaging data caused by for instance hardware or software
malfunction, participant’s movement inside the scanner or metallic objects on the
participant's person. Quality control was done using FSLView
(http://fsl.fmrib.ox.ac.uk/fsl/fslview/; Jenkinson, Beckmann, Behrens, Woolrich &
Smith, 2012), MRIcro (www.mricro.com; Rorden & Brett, 2000) and Statistical
Parametric Mapping Software 12 (SPM12; Wellcome Department of Cognitive
Neurology, London, UK) implemented in MATLAB (Mathworks, Natick, MA,
USA). Preprocessing of the data was done by initially realigning all functional
volumes to the mean image for each run, whereafter slice-timing correction was
performed. Slice-timing correction is done to correct for variability in the
measured BOLD-response caused by slices from the same volume being acquired
at different time points, as the first and last slice of a volume are acquired almost
one repetition time, in this case 2000ms, apart. Further, functional images (2 x 2 x
2 mm3 voxel dimensions) were coregistered to the structural images for each
individual, whereafter they were normalized by warping the images onto the
Montreal Neurological Institute (MNI152) template. These steps were taken to
facilitate anatomical identification of event-related activation. The images were
then smoothed using an isotropic 8-mm full-width-half maximum gaussian kernel
so as to reduce signal detection of non-systematic high-frequency spatial noise.
Data analysis
All analyses using non-imaging data were performed using SPSS statistics, v. 23.0
(IBM, Armonk, NY, USA). For all regression analyses, the outcome measure was
defined as residual gain scores on the LSAS-SR between first baseline and
posttreatment measurement points. Residual gain scores are a standardised
measure of change between two points in time. They were computed by first
standardising the raw scores on LSAS-SR to Z-scores, and then calculating the
difference between the posttreatment score, and the baseline score multiplied by
the correlation between the two (as such: Z2 - Z1*r12) as described by Manning and
Page 29
22
Dubois (1962). By factoring in the correlation between the two scores, initial
differences between participants are controlled for as well as measurement error
resulting from repeated measures on the same instrument (Steketee & Chambless,
1992), making it uniquely suited for measuring the correlates of change. Further,
controlling for measurement error due to repeated measures is of high relevance in
the context of the present study as participants were required to fill out the LSAS-
SR a total of 15 times across the duration of the study.
Response times from the experimental paradigm were analysed by aggregating
data and calculating a mean for each individual for each condition. The mean of
means within each condition were then compared using a paired sample t-test. All
reaction times below 300ms were treated as missing data and excluded from the
calculation of the individual means, as it is unlikely that individuals were able to
read and comprehend the sentences during that time. If the participants failed to
respond within 2500ms the response was not registered and treated as a failed trial,
thus excluding that particular response time from the calculation of individual
means. If individuals failed to respond in all cases of one condition they were
excluded from all subsequent analyses.
Behavioural measures predicting outcome
To assess the effect of pretreatment SAD symptom severity on treatment response,
LSAS-SR scores from pretreatment were correlated with the difference between
LSAS-SR scores at posttreatment and pretreatment. As earlier noted, subsequent
analyses were performed with residual gain scores of the LSAS-SR as the main
outcome variable, and the hypothesised predictors of depressive symptom severity,
according to MADRS-S, and treatment credibility/expectancy, measured using the
modified version of the CEQ described above, were entered as regressors in a
multiple regression analysis. To control for possible effects of age and gender,
these variables were entered in the first step of the regression model, prior to
entering the supposed predictors. Collinearity diagnostics were assessed to check
for possible multicollinearity.
Analysis of imaging data
The fMRI data was analysed using SPM12 and fitting the BOLD activation to the
voxel-wise general linear model (GLM). In the first-level analysis, the within-
subject level, two contrasts were modelled contrasting targets, self- and other-
referential criticism, to each other. In this step the movement parameters, extracted
during preprocessing, were included and controlled for as covariates of no interest.
The movement parameters consist of the individual's head movement in relation to
the mean image for each run, made up of six regressors: three for translation, and
Page 30
23
three for rotation. This was done to further control for motion artefacts. The
contrasts, which consist of differences in BOLD activation, were modelled as
following: other-referential criticism > self-referential criticism (Other > Self) and
self-referential criticism > other-referential criticism (Self > Other). Additionally,
the respective targets were contrasted to an implicit baseline (Self > Fixation
Cross; Other > Fixation Cross), but these were not analysed in the current thesis as
the primary interested was in the differential processing of self- and other related
criticism. The conditions were modelled as box-car functions, convolved with the
hemodynamic response function using a 128 s high-pass filter. As such the first-
level constitutes the computation of activation for each subject in each voxel, and
in the second-level a group map is formed from the first-level consisting of all
participants and computing one t-value for each voxel, representing activation
(Ashby, 2011). The second-level analysis is simply the GLM estimation of
specified contrasts over all participants. Exploratory whole-brain analyses without
covariates of interest were performed to explore activation related to each contrast.
This was done in order to investigate the differential activation related to the
various contrasts and thus evaluate whether the paradigm produces activation
patterns related to a proposed core mechanism of SAD. All results from contrast
analyses reported below are reported with a significance level of p < 0.001,
uncorrected, as this is common practice.
Analysing the predictive value of the imaging data was done by first performing a
whole-brain analysis of the Other > Self, and Self > Other contrasts including
residual gain scores as a covariate of interest, allowing us to identify which
differential activations were correlated with treatment gains. Beta-weights were
extracted from peak voxels in regions identified in the previous analysis and
subsequently used as continuous predictors in a regression model, so as to identify
the explained variance accounted for by the brain measures. Behavioural measures
and demographic characteristics were then entered in a second step, so as to see
how much variance these added, over and above the brain measures. Finally,
whole-brain analyses were once again performed with the behavioural measures
stepwise added as covariates of no interest, thus controlling for the variation in
brain activity related to each behavioural measure, so as to identify which brain
activity was uniquely related to treatment gains irrespective of the behavioural
measures.
Page 31
24
Results
Exploratory analyses of activation related to self- and other-referential criticism
As previously described participants were exposed to sentences of self-referential
and other-referential criticism and were required to press a button once they had
finished reading the sentence. Across participants, 1648 sentences were
administered for each target. There was a significant difference in the average
response time for other-referential (M = 1388ms, SD = 372ms; Mdn = 1407ms) and
self-referential (M = 1324ms, SD = 360ms, Mdn = 1350ms) sentences, t(44) =
5.254, p < 0.001, indicating a faster response time for self-referential sentences.
Two participants were excluded from the analyses due to not responding to any
sentence in one of the conditions within 2500ms.
In response to Other > Self, the participants showed enhanced activity in clusters
containing the dPCC, right occipital gyrus, bilateral angular gyrus, left middle
temporal gyrus, bilateral supplementary motor area (SMA), left vmPFC and left
dmPFC, see Figure 3 (for a complete list of statistically significant activations see
Table F1 in Appendix F). Furthermore significant peak level activity was observed
in the left subgenual cingulate cortex (sACC). In response to Self > Other,
enhanced activity was observed in bilateral supramarginal gyrus, right dPCC, right
postcentral gyrus, right occipitotemporal area, right posterior middle temporal
gyrus and left hippocampus see Figure 4 (for a complete list of statistically
significant activations see Table F1 in Appendix F). Notably, little differential
Figure 3 Figure 4
Other > Self. All shown clusters significant at p
< 0.001 (uncorrected). Sagittal view, x = -5,
Coronal view, y = -58, Axial view, z = 24
Self > Other. All shown clusters significant at
p < 0.001 (uncorrected). Sagittal view, x = -48,
Coronal view, y = -28, Axial view, z = 24
Page 32
25
activation was found in the previously outlined fear circuitry, for example the
amygdalae and insulae. To ensure that the paradigm indeed elicited anxiety-related
activation, inspection of the Other > Fixation Cross and Self > Fixation Cross
contrasts was performed and fear circuitry activation was found in both cases
(illustrated in Appendix G, Figures G1 and G2 respectively).
Predicting treatment response using behavioural measures and demographic
characteristics
The mean LSAS-SR score at pretreatment, nine weeks prior to treatment onset,
was 76.0 (SD = 19.0) and the mean score at posttreatment was 48.9 (SD = 24.1),
representing a statistically significant average decrease of 27.1 points (SD = 22.0,
t(45) = 8.37, p < .001, Cohen’s d = 1.23). The pretreatment LSAS-SR score was
negatively and statistically significantly correlated with change scores calculated as
LSASpost - LSASpre (r = -.31, p = .03), suggesting that higher pretreatment
symptom severity was related to a greater decrease in LSAS-SR scores between
pretreatment and posttreatment. The residual gain scores calculated from the
LSAS-SR were regressed upon age and gender in a first step, followed by
MADRS-S scores (M = 13.1, SD = 6.2) and CEQ scores (M = 31.7, SD = 8.3) in a
second step as previously described. Results from the ensuing models are reported
in Table 2. The first model containing age and gender was nonsignificant (F(2, 43)
= 1.778, p = .181), the second however was significant (F(4, 41) = 5.073, p <
0.001), explaining 27% of the variance. CEQ-scores emerged as the only
significant predictor. As age appeared to be largely unrelated to residual gain
Table 2
Summary of regression analysis predicting residual gain scores on the LSAS-SR from age,
gender, depressive symptom severity and treatment expectancy.
R2 R
2adj B SE B β t p
Step 1 0.08 0.03
Constant -1.02 0.75 - -1.37 0.179
Age 0.01 0.02 0.06 0.40 0.693
Gender 0.51 0.27 0.29 1.89 0.067
Step 2 0.33 0.27
Constant 0.38 0.79 - 0.49 0.628
Age 0.00 0.01 0.01 0.08 0.933
Gender 0.45 0.24 0.25 1.86 0.070
MADRS-S 0.03 0.02 0.25 1.87 0.069
CEQ -0.05 0.01 -0.48 -3.72 0.001
Note. ΔR2 = 0.255 for step 2 (p < 0.001), MADRS-S, Montgomery-Åsberg Depression
Rating Scale; CEQ, Credibility/Expectancy Questionnaire
Page 33
26
scores, it was excluded from subsequent analyses. Gender and MADRS-S were
however retained as they trended towards significance (p < 0.08). Collinearity
diagnostics revealed no issues regarding tolerance (all VIFs < 1.15; see Table H1
in Appendix H for correlational matrix).
Prediction analyses using brain responses
For the whole-brain analyses with
residual gain scores as a covariate of
interest, the greatest activation was
found along the midline structures
primarily in the left posterior mid
cingulate cortex (pMCC), left
anterior middle cingulate cortex
(aMCC), and the lingual gyrus (LG)
for the Other > Self contrast. The
only differential increased activation
in the Self > Other was a one voxel
cerebellum activation. The results are
reported in Table 3, and illustrated in
Figure 5.
Table 3
Whole-brain univariate voxel-wise test with residual gain scores for LSAS-SR as covariate of
interest
MNI-coordinates Volume
Contrast Region x y z (mm3) Z
Other > Self
L Posterior middle cingulate cortex -22 -20 44 416 3.79
L Anterior middle cingulate cortex -12 12 50 464 3.64
R Anterior middle cingulate cortex 14 16 44 176 3.57
L Lingual gyrus -14 -66 -2 288 3.51
L Primary visual cortex -12 -80 2 264 3.29
R Medial supplementary motor area 12 24 48 40 3.27
L Middle occipital gyrus -10 -96 -2 48 3.21
R Middle occipital gyrus 10 -74 0 32 3.18
Self > Other
L Cerebellum -14 -36 -20 8 3.45
Note. All listed clusters significant at p < 0.001 (uncorrected), MNI, Montreal Neurological
Institute; Z, Z-score.
Figure 5
Other > Self. All shown clusters significant at p <
0.001 (uncorrected). Sagittal view, x = -11,
Coronal view, y = -10, Axial view, z = 44
Page 34
27
Beta weights were extracted from peak voxels from two clusters showing
significant activation on the Other > Self contrast including residual gain scores as
a covariate of interest. The peak voxels chosen were from clusters in the left LG
(MNI Coordinates x = -14, y = -66, z = -2) and the pMCC (x = -22, y = -20, z =
44). These coordinates were chosen as they represent the two distinct areas in the
brain, the mid cingulate cortex and the inferior occipital lobule, in which activation
was shown to be correlated with residual gain scores (See Figures 6 and 7). Results
from the subsequent multiple regression analysis are reported in Table 4. In
summary, the brain measures by themselves explained 34% of the variance (as
measured by adjusted R2) and adding the behavioural measures and demographic
characteristics into the model resulted in the model explaining a total of 50% of the
variance. The brain measures were significant predictors in both models, at the p <
0.05 level, and CEQ and gender were also significant predictors in the final model.
Once again collinearity diagnostics revealed no issues regarding tolerance (all
VIFs < 1.35; see Table H1 in Appendix H for correlational matrix).
To further elucidate the relationship between the variance explained by the
behavioural measures and the variance explained by activity in the LG and the
pMCC, analyses were performed to discern which brain activation was related to
treatment response, but unrelated to the behavioural measures. This was done by
stepwise inserting gender, MADRS-S and CEQ, as covariates of no interest into
Table 4
Summary of regression analysis predicting residual gain scores on the LSAS-SR using beta-
weights from the Other > Self contrast, as well as including, gender, depressive symptom
severity and treatment expectancy in a second step
R2 R
2adj B SE B β t p
Step 1 0.37 0.34
Constant -0.03 0.10 - -0.26 0.798
L pMCC -1.66 0.57 -0.39 -2.94 0.005
L LG -0.48 0.19 -0.33 -2.51 0.016
Step 2 0.55 0.50
Constant 0.14 0,47 - -0.29 0.774
L pMCC -1.24 0.52 -0.29 -2.34 0.022
L LG -0.42 0.17 -0.29 -2.50 0.017
Gender 0.40 0.19 0.23 2.10 0.044
MADRS-S 0.03 0.02 0.19 1.67 0.102
CEQ -0.04 0.01 -0.35 -3.10 0.004
Note. ΔR2 = 0.18 for step 2 (p < 0.003). pMCC, Posterior Mid Cingulate Cortex; LG,
Lingual Gyrus; MADRS-S, Montgomery-Åsberg Depression Rating Scale; CEQ,
Credibility/Expectancy Questionnaire.
Page 35
28
the whole-brain analysis with residual gain scores as the covariate of interest.
Generally covarying for the above measures reduced statistically significant
activity somewhat, but no large differences were related to any one of the
measures. In short, covarying for gender decreased statistically significant activity
in the occipital cluster, covarying for MADRS-S increased the cluster size of the
smaller cluster in the left pMCC, and covarying for CEQ removed the cluster in the
left aMCC. Thus MADRS-S scores appeared to be related to suppressed
differential activation otherwise related to greater treatment gains while, being
female and having a higher score on the CEQ appeared to be related to greater
activation in areas related to treatment gains. Controlling for two at a time did not
substantially alter the above findings. Finally when controlling for all of the above
the participants showed activity in clusters within the left LG, right aMCC and left
pMCC, see Figure 8. For a complete list of statistically significant activations
when controlling for all covariates see Table 5.
Figure 6
Showing the relationship between BOLD
response from the left posterior mid cingulate
cortex (pMCC) and residual gain scores.
Coordinates according to MNI, Montreal
Neurological Institute: x = -22, y = -20, z =
44. Lower residual gain scores signifying
greater treatment response.
Figure 7
Showing the relationship between BOLD
response from the left lingual gyrus (LG) and
residual gain scores. Coordinates according to
MNI, Montreal Neurological Institute: x = -14,
y = -66, z = -2. Lower residual gain scores
signifying greater treatment response.
Page 36
29
Table 5
Whole-brain univariate voxel-wise test for Other > Self and Self > Other controlling for gender,
MADRS-S, CEQ with residual gain scores for LSAS-SR as dependent variable
MNI-
coordinates
Volume
Contrast Region x y z (mm3) Z
Other > Self
L Lingual gyrus -12 -60 -8 296 3.57
L Posterior middle cingulate cortex -22 -20 44 16 3.22
L Posterior middle cingulate cortex -14 -20 44 64 3.19
R Anterior middle cingulate cortex 14 16 42 8 3.19
R Superior temporal gyrus 42 -26 2 8 3.12
Self > Other
No significant activationsa
Note. All listed clusters significant at p < 0.001 (uncorrected), MNI, Montreal Neurological
Institute; Z, Z-score. a Artefact activation found in right lateral ventricle with 80mm
3 activation.
Figure 8
Other > Self controlling for gender, MADRS-S and
CEQ with residual gain scores as dependent
variable. All shown clusters significant at p < 0.001
(uncorrected). Sagittal view, x = -15, Coronal view,
y = -64, Axial view, z = -7
Page 37
30
Discussion
In the current study, fMRI was used to evaluate the BOLD-responses of 46
participants with SAD, to a disorder-relevant paradigm in which they were
subjected to negative statements directed either towards themselves or a significant
other. This was done in order to investigate the neural correlates of a central part of
SAD, the fear of negative evaluation, as well as to appraise the predictive value of
functional imaging data as regards treatment response to ICBT for SAD. To
summarise the results: the exploratory analyses revealed differential activation in
the Other > Self as well as the Self > Other contrasts, indicating that the different
conditions were related to differential processing. It was hypothesised that whole-
brain analyses would reveal differential activity situated in the fear circuitry, this
was however not observed. The large effect size observed for the ICBT treatment
(Cohen’s d = 1.23) was comparable to previous studies (Boettcher et al., 2013),
with the majority of participants showing improvement. It should once again be
noted that posttreatment scores were collected subsequent to treatment completion,
but prior to the final MRI-scan in the research project. Regarding behavioural
measures and demographic characteristics, these explained 27% of the variation in
treatment response. CEQ-score was the only predictor variable significantly related
to treatment response in the model assessing gender, age, MADRS-S score and
CEQ score and their relation to residual gain scores on the LSAS-SR. Analyses
relating BOLD activation to treatment response revealed activation on the Other >
Self contrast, with the primary clusters situated bilaterally in the aMCC, in the left
pMCC, the left LG extending towards the left primary visual cortex. Contrary to
what had been hypothesised, whole-brain analyses of the Other > Self and Self >
Other contrasts did not reveal any differential activation related to treatment
response in the amygdala, the insula, or the ACC. Beta-weights were extracted
from the cluster showing the greatest activation in the MCC, as well as the cluster
in the LG and subsequently used in a multiple regression. When modelled together
with the behavioural measures, the aggregated model explained 50% of the total
variance in outcome, supporting the stated hypothesis that the combination of brain
measures and behavioural measures would explain more variation than behavioural
measures alone. Finally, gender, MADRS-S and CEQ were included as covariates
of no interest in a whole brain analysis containing residual gain scores as the
covariate of interest. Results from these analyses showed that activity in the LG as
well as the left pMCC was related to treatment gains whilst controlling for the
effects of gender, MADRS-S and CEQ.
Differential responses to self- and other-referential criticism
The results from the paired sample t-test for response times suggest that when
viewing self-referential criticism participants responded with faster confirmation
Page 38
31
and subsequent removal of the stimuli. This could indicate that self-referential
criticism was more anxiety provoking and thus prompted a faster response time.
This is in line with findings reporting an increased avoidance of fear provoking
stimuli with increasing anxiety (Mogg, Bradley, Miles, & Dixon, 2004), something
which have also been illustrated in individuals with social anxiety (Heuer, Rinck,
& Becker, 2007). This interpretation is however not supported by the result on the
whole-brain analyses for Self > Other, and Other > Self contrasts, which displayed
no increased activation in the fear circuitry.
The exploratory analysis for the Other > Self contrast, without covarying for other
measures, revealed statistically significant activation in parts of the left occipital
gyrus, PCC, Angular gyri and dmPFC, areas which have previously been
implicated in the so called Default Mode Network (DMN; Raichle, 2015). The
DMN is a large network of brain areas, consisting roughly of vmPFC, dmPFC,
PCC, precuneus and lateral parietal cortex, that are active during resting or during
attention to internal states, and has been implicated in emotional processing, self-
referential mental activity and recollection of prior experiences (Raichle, 2015).
The DMN, usually studied via resting state fMRI, has previously been linked to
SAD (Peterson, Thome, Frewen & Lanius, 2014), and mixed results have been
presented with studies showing increased connectivity in the network (Liao et al.,
2011), no difference compared to healthy controls (Pannekoek et al., 2013), and
decreased connectivity in areas such as the PFC, the PCC and the AG (Doruyter et
al., 2016; Gentili et al., 2009; Qiu et al., 2011). A complementary view comes from
research on fear conditioning where it has been found that viewing a non-
conditioned stimulus, CS- also known as a safety signal activation is seen in these
very same areas, implicated in DMN (Fullana et al., 2016). As such, safety cues
provoke brain activation similar to when the brain is at rest, during internal focus
or recollection of prior experiences. In the current sample of participants with
SAD, greater DMN activation related to the other-referential stimuli suggesting
that they were interpreted as CS- relative to the self-referential stimuli.
Interestingly, Eisenberger et al. (2011) showed that being near an attachment figure
during a threatening experience produced activation in the vmPFC similar to that
related to safety signals. As many of the participants in the current sample likely
selected attachment figures as their significant other, it is quite possible that the
other-referential stimuli induced this sort of attachment-related safety signal,
explaining the activation pattern seen on the Other > Self contrast. The exploratory
analysis for the Self > Other contrast, without covarying for other measures,
showed significantly greater activation in the bilateral supramarginal gyrus and
right dPCC. The activation in the superior and inferior supramarginal gyrus area
overlaps almost entirely with the secondary somatosensory cortex (SII; Ruben et
Page 39
32
al., 2001). The SII has been strongly implicated in the processing of pain
(Apkarian, Bushnell, Treede & Zubieta, 2005). Furthermore a large overlap has
been demonstrated between the areas for physical- and social pain processing,
indicating the SII is central to the processing of social pain (Kross, Berman,
Mischel, Smith & Wager, 2011). Fullana et al. (2016), showed in their meta-
analysis that when contrasting conditioned feared stimuli, CS+, to a CS-, increased
activation was found in the SII, indicating this area to be related to fear processing.
Additionally when controlling for confounding variables the SII seemed to be more
involved in the anticipatory response of CS+. Relating to the current results this
could imply that even though we did not find a greater activation in the previously
outlined fear network for the Self > Other contrast, the activation in the SII could
yet indicate an increased fear response. As the region has also been implicated as
playing a central part in the sensory processing of pain, originating from physical
and social sources alike (Eisenberger, 2012) it is possible that the participants
experienced what would be described as social pain, and a related fear response.
The activation showing on the Self > Other contrast would thus indicate social
pain, related to the fear of negative evaluation central to SAD, producing a
differential fear response not in the areas most related to the fear circuitry, but in
the SII.
Predicting treatment response using behavioural measures and
demographic characteristics
Multiple regression analysis of behavioural measures hypothesised to predict
treatment response provided a statistically significant model, consisting of age,
gender, MADRS-S scores and CEQ scores, explaining 27% of the variance. CEQ
score at pretreatment emerged as the only significant predictor, with MADRS-S
score and gender trending towards significance. These behavioural measures, not
including age, were later modelled along with brain measures; the brain measures
by themselves explaining 34% of the variance, and the aggregated model
explaining 50% of the variance. Activity in both the pMCC and the LG emerged as
significant predictors in the final model. CEQ score and gender were also
significant predictors, such that higher treatment expectancy prior to treatment, and
being male, were related to greater treatment response. The effect of gender should
however be interpreted cautiously due to the skewed distribution, 63% of the
sample being women. Further, the current results illustrate that brain activations
from two areas in the brain added a great deal of value to the prediction of
treatment response in the current sample, pointing to the importance of further
studying neuroimaging as a way of predicting who will respond to treatment.
Notably however, behavioural and demographic predictors were chosen a priori,
while the brain measures were chosen post hoc, due to being the two most
Page 40
33
predictive areas in the current sample. Assuming the purpose of using
neuroimaging methods in prediction research should be to identify patterns of
neural activity which can be used to reliably predict who will respond to what
treatment.
The current paradigm is supposedly related to the fear of negative evaluation,
future studies could evaluate paradigms related to other proposed core processes in
SAD. As an example, Mörtberg and Andersson (2014) studied the predictive value
of pretreatment fear of negative evaluation, as well as the predictive value of two
character dimensions, namely anxious worry and self-directedness. The authors
found fear of negative evaluation to be predictive of residual gain scores at
posttreatment in individual cognitive therapy, but not for participants undergoing a
similar group therapy. Anxious worry at pretreatment however was negatively
related to residual gain scores at posttreatment as well as 1-year follow-up for the
entire sample, regardless of treatment received. As such, fMRI-paradigms which
elicit activation related to the construct of anxious worry could be investigated for
their predictive value CBT for SAD. In general, neuroimaging research aiming to
predict treatment response would likely benefit from understanding the neural
correlates of behavioural measures which have previously been shown to predict
treatment response, and developing experimental tasks by which these activations
can be evaluated.
As previously mentioned, results from the multiple regression model of possible
behavioural predictors revealed CEQ scores as the only significant predictor, such
that higher CEQ scores were related to greater treatment gains. Treatment
credibility and treatment expectancy, as measured by the CEQ, this appears to be
emerging as one of few stable predictors of treatment response in CBT for SAD. It
has been proposed that the credibility index measures cognitive appraisal of the
treatment, while the expectancy index measures affective appraisal (Devilly &
Borkovec, 2000), and that these therefore measure two aspects of the same
underlying construct (Boettcher, Renneberg & Berger, 2013). Further, it has been
suggested that treatment expectancy and treatment credibility are related to
additional processes important for psychotherapy outcome, such as self-efficacy
and readiness for change (Price & Andersson, 2012). Boettcher, Renneberg and
Berger (2013) found that part of the effect treatment expectancy had on treatment
outcome was mediated by early treatment gains, and by improved adherence, but
also found that treatment expectancy asserted a direct effect beyond that explained
by the mediators. Results such as these highlight the importance of further
investigating through which processes treatment expectancy affects treatment
response to enable further development of treatment methods. For a long time,
Page 41
34
treatment expectancy was regarded as a variable to be factored out, so as to isolate
effects unrelated to expectation effects (Constantino, Arnkoff, Glass, Ametrano &
Smith, 2011). Constantinto et al. however discuss the fact that expectations about
treatment in recent years have been acknowledged as a central process in
psychological treatment, and an important consideration in the choice of treatment
to be delivered, as well as for the therapist to actively cultivate treatment
credibility by providing a clear rationale, and positive treatment expectancy by
acknowledging treatment success and reinforcing the patient's self-efficacy relating
to their ability to change. As an example, Nordgreen et al. (2012) found that
treatment expectancy was related to treatment adherence in unguided, but not in
guided, self-help for SAD, perhaps alluding to an effect of therapist guidance in
mitigating early concerns from the patients. Regarding the current sample,
expectancy and credibility measures were likely affected by the fact that
participants applied for the study of their own volition, expecting to receive an
evidence based, internet-delivered treatment for their SAD. The participants also
underwent a comprehensive screening process, involving multiple contacts with
research personnel, as well as attending two MRI-sessions prior to commencing
treatment. It is feasible that this affected the participants’ ratings of credibility and
expectancy, as they formed an opinion on the competency and credibility of the
research staff and the project in general. The context of the study therefore
produces a stark contrast to a hypothetical naturalistic setting, where patients,
unfamiliar with common concepts of psychological treatment, seek treatment for
SAD through outpatient psychiatric care and are offered ICBT as their only
alternative. In such a setting, attention to the expectations that patients have on the
treatment can be of great importance to maximise treatment response. This has also
been discussed in the previously mentioned study by El Alaoui, Ljótsson et al.
(2015) who investigated the effect of treatment credibility in an outpatient ICBT
treatment for SAD and found treatment credibility, measured at the second week of
treatment, to be the strongest prognostic factor of both adherence and symptomatic
improvement. The authors noted that their participants specifically applied for
internet-delivered treatment, and that it would be interesting to compare the
predictive value of treatment credibility in a sample where patients had varying
preferences regarding the modality of treatment delivery.
Prediction analyses correlating brain responses to residual gain scores
Results from the current analyses assessing the predictive value of brain activation
indicate activity in the anterior and posterior MCC as well as the LG at nine weeks
prior to treatment as being predictive of treatment outcome. Regarding the aMCC
and the pMCC, these areas have been shown to be activated by a multitude of tasks
(Vogt, 2005; Vogt, Berger & Derbyshire, 2003), with the aMCC, commonly
Page 42
35
misnamed as the dACC (Vogt, 2016), involved in context evaluation as well as
regulation of behavioural strategies including pain avoidance and reward
sensitivity, and pMCC involved in spatial orientation of the own body along with
pain processing and attention to sensory stimuli including harmful stimuli (Vogt,
2016). Regarding social situations, this is resonated in Apps, Lockwood, and
Balsters, (2013) who illustrate that the MCC is involved in predicting and
monitoring outcome during social interaction. Moreover, the aMCC has been
implicated in the perception (Müller et al., 2010) and regulation of aversive social
stimuli (Koenigsberg et al., 2010) and the pMCC also having been found to be
activated in the perception of social stimuli (Gaebler et al., 2014) but additionally
facilitating cognitive appraisal as an emotion regulation technique (Diekhof, Geier,
Falkai & Gruber, 2011). Relating to our results, where greater activation on the
Other > Self contrast was related to greater treatment outcome, these findings
regarding the aMCC and pMCC could suggest that individuals who do better in
treatment have a greater ability to regulate their negative emotions and differentiate
between stimuli relating to the self or others. Interestingly, the cingulate cortex is
one of the most studied areas as pertains prediction of treatment response for both
the pharmacological and psychological treatment of psychiatric disorders. Activity
in the cingulate cortex has been shown to predict treatment response to CBT for
SAD (Klumpp, Fitzgerald & Phan, 2013; Månsson et al., 2015), Major Depressive
Disorder (Siegle, Carter & Thase, 2006; Fu et al., 2008), Panic Disorder (Lueken et
al., 2013) and Generalized Anxiety Disorder (Nitschke et al., 2009). In general
however, the activity found has been in the ACC, the rostral part of the cingulum,
which has been found to be involved in fear extinction (Sehlmeyer et al., 2009).
One reason for the differential findings currently reported could be the use of an
experimental task which is more related to cognitive components of SAD, and
therefore activates areas more related to cognitive reappraisal, rather than fear
extinction. Månsson et al. (2015) however investigated a previous version of the
paradigm using MVPA and found activity in the dACC and amygdala to be highly
predictive of treatment response.
Brühl et al. (2014) argued that one of the foremost conclusions of their meta-
analysis was the inclusion of the posterior parietal and occipital areas into the
neurobiological explanation for SAD which had prior to this been neglected. Brühl
and colleagues proposed, based on their analysis, that the occipital cortex showed
enhanced activity, and increased connectivity to the frontal lobule and amygdalae
as well as decreased connectivity to the PCC. In the current study, results suggest a
relationship between activity in the LG on the Other > Self contrast relating to
treatment outcome. Activation in occipital areas at pretreatment have previously
been related to treatment outcome in CBT for SAD, but only when viewing
Page 43
36
emotional faces, and not including the LG (Doehrmann et al., 2013; Klumpp,
Fitzgerald & Phan, 2013), suggesting different underlying processes as compared
to the current results. Even though the LG is extensively involved in visual
processing (Kastner & Ungerleider, 2000), it has also been suggested to influence
cognition as well as emotion (Pessoa, 2008) and being involved in shaping
perception depending on emotional states and stimuli (Lindquist, Wager, Kober,
Bliss-Moreau & Barrett, 2012). The LG has been shown to be more active during
fear learning (Damaraju, Huang, Barrett & Pessoa, 2009), memory of a fear
conditioned stimulus (Moratti & Keil, 2009), as well as viewing threatening stimuli
such as threat words (Isenberg et al., 1999) and angry faces (Morris, Buchel &
Dolan, 2001), thus LG seems to be involved in fear processing. Results from the
current study showed that differential activation in the LG depending on the target
of negative social stimuli was related to residual gain scores, further indicating a
possible association between aberrant fear processing and treatment response in
SAD. It should be noted that because of the connectivity network of the LG, the
emotional valence detection in visual stimuli could be driven by backward
feedback from the amygdala (Pessoa, 2008), even though the processing itself is
done by the visual cortex involving the LG (Kastner & Ungerleider, 2000). Thus
our results seem to be in line with Brühl et al. (2014), stating that the occipital
regions presumably have integrating and regulating functions in regard to emotion,
and that these are irregular in people with SAD.
To summarise the above, activity in fear circuitry areas a priori hypothesised to be
related to treatment did not emerge as significant predictors of treatment response
on whole-brain analyses for neither the Other > Self nor the Self > Other contrasts.
The results did however indicate that treatment response was predicted by activity
in an area involved in fear processing of emotionally valenced visual stimuli, the
LG, an area associated with using cognitive reappraisal to regulate emotion, the
pMCC, as well as an area related to minimizing pain and maximizing reward,
aMCC. This is not entirely surprising as cognitive reappraisal and being able to
tolerate discomfort in order to achieve long-term goals are core components of the
CBT-treatment used in this study. Moreover, aberrant visual processing and
hypervigilance to social threat could possibly be risk factors of slower rate of
change. As an example Jung et al. (2014) found differential volume of the LG to be
related to neuropsychological deficits and subsequent reduced response to
pharmacological treatment for Major Depressive Disorder. Fu et al. (2008) also
investigated participants with Major Depressive Disorder, and analysing their
cerebral responses to viewing sad faces found activation in the LG as well as the
cingulate cortex to be predictive of treatment response to CBT, further indicating
that these areas are important to achieve change during CBT-treatment. It is
Page 44
37
interesting to note that a previous study by Frick, Howner, Fischer, Eskildsen et al.
(2013) showed increased cortical thickness in the LG for participants with SAD
relative to healthy controls, while the current study indicates a role for this area in
predicting treatment response, suggesting that further research specifically target
this area.
Limitations
There are several limitations to be noted. First, the representativity of the current
sample can be called into question, not because of demographic characteristics, but
because of the fact that despite their social fears, these individuals were willing to
undergo a rigorous screening process, involving hours of online screening, several
phone calls in which they were diagnostically interviewed, and finally appear at an
MRI-lab, where they allowed unfamiliar professionals and researchers to lock them
inside an MRI-camera for an hour. Yet these individuals consented to go through
this process, four times, knowing what it would entail. Notwithstanding the awe
and gratitude the current authors feel towards these participants, it is safe to say
that not everyone suffering from SAD would have done the same. It might be
suggested that these participants were highly motivated to change at onset,
reflected by the large overall treatment response, and that they had an ability
uniquely suited for CBT treatment, namely the ability to willingly endure
discomfort in the service of long-term gains.
Second, in the current paradigm criticism related to both a significant other and the
self are likely to elicit social fear, thus when contrasting the conditions against
each other activation related to social fear is cancelled out, if not significantly
different depending on the target. Notably it has been shown that viewing a
significant other receiving negative comments elicits activation possibly related to
an empathic social response, indicating a shared social pain (Bernhardt, & Singer,
2012). This could possibly complicate the interpretation of the current results, as it
is uncertain whether the activation related to other-referential criticism in the
current paradigm elicits similar social fear to that of self-referential criticism, and
specifically how this is related to SAD. As such, including a control group could
have possibly shed additional light on how individuals with SAD experience social
pain differently than healthy controls. Also, multilevel analyses of the response
times could have been used to account for the nested data structure and repeated
measures, so as to further elucidate the nature of the difference between the
conditions.
Third, the brain regions a priori hypothesised to be related to treatment outcome
did not yield any significant results when performing whole-brain analyses, why
Page 45
38
data was analysed to specifically identify regions related to residual gain scores.
This kind of procedure, where data are unselectively explored to identify
correlations between brain activation at pretreatment and treatment response,
statistically significant at an arbitrary threshold, and uncorrected for multiple
comparisons, is especially problematic in neuroscience, where over a hundred
thousand comparisons are done for every analysis. Further, no analyses of
connectivity were perfosrmed, restricting our results merely to differential
activation of particular voxels, and no minimum cluster size was specified,
increasing the risk of random voxels exceeding the significance threshold. The
ensuing results must be regarded cautiously, knowing that they hold true only for
the current sample, for the specific paradigm, and that knowledge about how
generalizable the results are can only be gained by replicating them in a different
sample. A way to counteract these problems regarding exploration of correlations,
would have been to use cross-validation measures such as leave-one-out or leave-
k-out, as it would prevent overfitting the models and promote generalisability
(Kriegeskorte, Simmons, Bellgowan & Baker, 2009). Another way to partially
bypass these limitations would have been to use a multivariate method such as the
previously mentioned MVPA-methods, which are currently gaining in popularity
among neuroscientist.
Fourth, the within-group design of the study, and consequent lack of a control
group, limits the interpretations possible from the current results. It would have
been of great interest to study the differences between the present sample and a
group of healthy controls on the contrasts set up for the paradigm used. As it was
assumed that people with SAD would react more negatively to reading criticism
directed at them, a control group would have allowed the validation of this
assumption. Furthermore, as discussed, our findings suggest that the participants
experienced the other-related criticism as a form of control-condition, or even a
safety signal, and it would have been interesting to gauge the reaction of healthy
control.
Future research
The current study was the first to apply the experimental task used in its current
form. Though our results suggest that using the name of a significant other in the
paradigm could possibly be interpreted by the participants as a safety cue, further
research should look to validate these findings by employing a control group to
investigate how the activation found in our sample relates to that of healthy
controls. Furthermore the areas found to be related to treatment response in this
study could be employed as regions of interests in future studies, thus validating
their relevance to prediction of treatment response. It would undeniably be of
Page 46
39
interest to also employ a fear conditioning paradigm to see the differential
activation to CS+ and CS- in this comparison. It has also been reported that
individuals with SAD have been shown to have elevated fear conditioning
responses to social stimuli (Lissek et al., 2008) and show resistance to extinction
(Hermann, Ofer & Flor, 2004), thus a fear conditioning paradigm might also be of
interest in predicting treatment outcome by gauging the individual's pretreatment
conditioning pattern.
Finally, research on predictors should strive to integrate behavioural measures and
neuroimaging methods, using one to further the other, by investigating the neural
correlates of behavioural predictors, such as treatment expectancy, and the
behavioural correlates of neural predictors.
Conclusions
Despite the limitations delineated above, certain conclusions can be drawn.
Exploring the neural reactions to significant other- and self-referential statements,
it was illustrated that the participants showed a strong differential activation for
each condition, with self-referential statements appearing to be more related to fear
processing. Furthermore, differential activation in the LG and the MCC predicted
treatment response, and combined with behavioural measures explained 50% of the
variance, once again illustrating the value of neuroimaging methods in prediction
research. As few behavioural measures have proven to be reliably predictive of
treatment response in CBT for SAD, it would be preferable to increasingly apply
neuroimaging to assess which patients would benefit from treatment. As this is
rather costly, it is however unlikely to be realistic considering the neuroimaging
methods currently available, and lacking a clear-cut cost-benefit analysis. Another
direction would be to further investigate neural correlates of treatment response, as
well as the behavioural output related to those neural responses, so as to enable the
development of behavioural measures likely to capture the same explained
variance. To conclude, ICBT appears to be an effective treatment for many, but not
all, people suffering from SAD. Being able to identify those individuals would
allow for a more cost-efficient psychiatric care, as well as enabling the further
development of treatment options for treatment-resistant SAD.
Page 47
40
References
Abraham, A., Kaufmann, C., Redlich, R., Hermann, A., Stark, R., Stevens, S., &
Hermann, C. (2013). Self-referential and anxiety-relevant information
processing in subclinical social anxiety: An fMRI study. Brain Imaging and
Behavior, 7(1), 35–48. doi:10.1007/s11682-012-9188-x
Acarturk, C., Smit, F., de Graaf, R., van Straten, A., ten Have, M., & Cuijpers, P.
(2009). Economic costs of social phobia: A population-based study. Journal
of Affective Disorders, 115(3), 421–429. doi:10.1016/j.jad.2008.10.008
Amaro, E., & Barker, G. J. (2006). Study design in fMRI: Basic principles. Brain
and Cognition, 60(3), 220–232. doi:10.1016/j.bandc.2005.11.009
American Psychiatric Association. (2000). Diagnostic and statistical manual of
mental disorders (4th ed., text rev.). Washington, DC: Author.
American Psychiatric Association. (2013a). Diagnostic and statistical manual of
mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.
American Psychiatric Association. (2013b). Diagnostic and Statistical Manual of
Mental Disorders. doi:10.1176/appi.books.9780890425596.dsm05
Andersson, G., Bergström, J., Buhrman, M., Carlbring, P., Holländare, F., Kaldo,
V., … Waara, J. (2008). Development of a New Approach to Guided Self-
Help via the Internet: The Swedish Experience. Journal of Technology in
Human Services, 26(2-4), 161–181.
http://doi.org/10.1080/15228830802094627
Andersson, G., Carlbring, P., Furmark, T., & SOFIE Research Group. (2012).
Therapist experience and knowledge acquisition in internet-delivered CBT for
social anxiety disorder: a randomized controlled trial. PloS one, 7(5), e37411.
Andersson, G., Carlbring, P., Holmström, A., Sparthan, E., Furmark, T., Nilsson-
Ihrfelt, E., Buhrman, M., & Ekselius, L. (2006). Internet-based self-help with
therapist feedback and in vivo group exposure for social phobia: a randomized
controlled trial. Journal of Consulting and Clinical Psychology, 74(4), 677–
686. http://doi.org/10.1037/0022-006X.74.4.677
Page 48
41
Andersson, G., Carlbring, P., Ljótsson, B., & Hedman, E. (2013). Guided internet-
based CBT for common mental disorders. Journal of Contemporary
Psychotherapy, 43(4), 223–233. http://doi.org/10.1007/s10879-013-9237-9
Andersson, G., Cuijpers, P., Carlbring, P., Riper, H., & Hedman, E. (2014). Guided
Internet‐based vs. face‐to‐face cognitive behavior therapy for psychiatric and
somatic disorders: a systematic review and meta‐analysis. World Psychiatry,
13(3), 288-295.
Andersson, E., Rück, C., Lavebratt, C., Hedman, E., Schalling, M., Lindefors, N.,
… Furmark, T. (2013). Genetic polymorphisms in monoamine systems and
outcome of cognitive behavior therapy for social anxiety disorder. PLoS ONE,
8(11). doi:10.1371/journal.pone.0079015
Apkarian, A. V., Bushnell, M. C., Treede, R. D., & Zubieta, J. K. (2005). Human
brain mechanisms of pain perception and regulation in health and disease.
European Journal of Pain, 9(4), 463–484. doi:10.1016/j.ejpain.2004.11.001
Apps, M. A. J., Lockwood, P. L., & Balsters, J. H. (2013). The role of the
midcingulate cortex in monitoring others’ decisions. Frontiers in
Neuroscience, 7(7 DEC), 1–7. doi:10.3389/fnins.2013.00251
Arnberg, F. K., Linton, S. J., Hultcrantz, M., Heintz, E., & Jonsson, U. (2014).
Internet-delivered psychological treatments for mood and anxiety disorders: A
systematic review of their efficacy, safety, and cost-effectiveness. PLoS ONE,
9(5). http://doi.org/10.1371/journal.pone.0098118
Ashby, F. G. (2011). Statistical Analysis of fMRI Data. Cambridge, US: MIT Press
Baker, S. L., Heinrichs, N., Kim, H. J., & Hofmann, S. G. (2002). The Liebowitz
social anxiety scale as a self-report instrument: A preliminary psychometric
analysis. Behaviour Research and Therapy, 40(6), 701–715.
doi:10.1016/S0005-7967(01)00060-2
Ball, T. M., Stein, M. B., Ramsawh, H. J., Campbell-Sills, L., & Paulus, M. P.
(2014). Single-subject anxiety treatment outcome prediction using functional
neuroimaging. Neuropsychopharmacology, 39(5), 1254–61.
http://doi.org/10.1038/npp.2013.328
Page 49
42
Barber, J. P. (2007). Issues and findings in investigating predictors of
psychotherapy outcome: Introduction to the special section. Psychotherapy
Research, 17(2), 131–136. http://doi.org/10.1080/10503300601175545
Barrera, T. L., & Norton, P. J. (2009). Quality of life impairment in generalized
anxiety disorder, social phobia, and panic disorder. Journal of Anxiety
Disorders, 23(8), 1086–1090. doi:10.1016/j.janxdis.2009.07.011
Beard, C., Moitra, E., Weisberg, R. B., & Keller, M. B. (2010). Characteristics and
predictors of social phobia course in a longitudinal study of primary-care
patients. Depression and Anxiety, 27(9), 839–845. doi:10.1002/da.20676
Bell, C. J., Malizia, A. L., & Nutt, D. J. (1999). The neurobiology of social phobia.
European Archives of Psychiatry and Clinical Neuroscience, 249(SUPPL.),
11–19. doi:10.1007/PL00014162
Bernhardt, B. C., & Singer, T. (2012). The Neural Basis of Empathy. Annual
Review of Neuroscience, 35(1), 1–23. doi:10.1146/annurev-neuro-062111-
150536
Blair, K. S., Geraci, M., Hollon, N., Otero, M., DeVido, J., Majestic, C., … Pine,
D. S. (2010). Social norm processing in adult social phobia: Atypically
increased ventromedial frontal cortex responsiveness to unintentional
(embarrassing) transgressions. American Journal of Psychiatry, 167(12),
1526–1532. doi:10.1176/appi.ajp.2010.09121797
Blair, K. S., Geraci, M., Otero, M., Majestic, C., Odenheimer, S., Jacobs, M., …
Pine, D. S. (2011). Atypical modulation of medial prefrontal cortex to self-
referential comments in generalized social phobia. Psychiatry Research -
Neuroimaging, 193(1), 38–45. doi:10.1016/j.pscychresns.2010.12.016
Blair, K., Geraci, M., Devido, J., Mccaffrey, D., Chen, G., Vythilingam, M., …
Pine, D. S. (2008). Neural response to self- and other referential praise and
criticism in generalized social phobia. Archives of General Psychiatry, 65(10),
1176–1184. doi:10.1001/archpsyc.65.10.1176.Neural
Page 50
43
Blair, K., Shaywitz, J., Smith, B. W., Rhodes, R., Geraci, M., Jones, M., … Pine,
D. S. (2008). Response to emotional expressions in generalized social phobia
and generalized anxiety disorder: evidence for separate disorders. American
Journal of Psychiatry, 165(9), 1193–1202.
doi:10.1176/appi.ajp.2008.07071060
Blair, K.S., Geraci, M., Korelitz, K., Otero, M., Towbin, K., Ernst, M., Leibenluft,
E., Blair, R.J., Pine, D.S., (2011). The pathology of social phobia is
independent of developmental changes in face processing. American Journal
of Psychiatry, 168, 1202–1209.
Boehme, S., Ritter, V., Tefikow, S., Stangier, U., Strauss, B., Miltner, W. H. R., &
Straube, T. (2013). Brain activation during anticipatory anxiety in social
anxiety disorder. Social Cognitive and Affective Neuroscience, 9(9), 1413–
1418. doi:10.1093/scan/nst129
Boehme, S., Miltner, W. H. R., & Straube, T. (2015). Neural correlates of self-
focused attention in social anxiety. Social Cognitive and Affective
Neuroscience, 10(6), 856–62. doi:10.1093/scan/nsu128
Boettcher, J., Carlbring, P., Renneberg, B., & Berger, T. (2013). Internet-based
interventions for social anxiety disorder - An overview. Verhaltenstherapie,
23(3), 160–168. http://doi.org/10.1159/000354747
Boettcher, J., Renneberg, B., & Berger, T. (2013). Patient expectations in internet-
based self-help for social anxiety. Cognitive Behaviour Therapy, 42(3), 203–
14. http://doi.org/10.1080/16506073.2012.759615
Borge, F. M., Hoffart, A., & Sexton, H. (2010). Predictors of outcome in
residential cognitive and interpersonal treatment for social phobia: Do
cognitive and social dysfunction moderate treatment outcome?. Journal of
behavior therapy and experimental psychiatry, 41(3), 212-219.
Borkovec, T. D., & Nau, S. D. (1972). Credibility of analogue therapy rationales.
Journal of Behavior Therapy and Experimental Psychiatry, 3(4), 257–260.
http://doi.org/10.1016/0005-7916(72)90045-6
Broca, P. (1865). Sur le siège de la faculté du langage articulé. Bulletins de la
Société d'Anthropologie de Paris, 6(1), 377-393.
Page 51
44
Brooks, S. J., & Stein, D. J. (2015). A systematic review of the neural bases of
psychotherapy for anxiety and related disorders. Dialogues in clinical
neuroscience, 17(3), 261.
Brühl, A. B., Delsignore, A., Komossa, K., & Weidt, S. (2014). Neuroimaging in
Social Anxiety Disorder–a meta-analytic review resulting in a new
neurofunctional model. Neuroscience & Biobehavioral Reviews, 47, 260–280.
doi:10.1016/j.neubiorev.2014.08.003
Brühl, A. B., Herwig, U., Delsignore, A., Jäncke, L., & Rufer, M. (2013). General
emotion processing in social anxiety disorder: Neural issues of cognitive
control. Psychiatry Research - Neuroimaging, 212(2), 108–115.
doi:10.1016/j.pscychresns.2012.05.006
Brühl, A. B., Rufer, M., Delsignore, A., Kaffenberger, T., Jäncke, L., & Herwig,
U. (2011). Neural correlates of altered general emotion processing in social
anxiety disorder. Brain Research, 1378, 72–83.
doi:10.1016/j.brainres.2010.12.084
Cahn, B. R., & Polich, J. (2006). Meditation states and traits: EEG, ERP, and
neuroimaging studies. Psychological Bulletin, 132(2), 180–211.
doi:10.1037/0033-2909.132.2.180
Carlbring, P., Gunnarsdóttir, M., & Hedensjö, L., Andersson, G., Ekselius, L., &
Furmark, T. (2007). Treatment of social phobia from a distance: A
randomized trial of internet delivered cognitive behavior therapy (CBT) and
telephone support. British Journal of Psychiatry, 190, 123-128. doi:
10.1192/bjp.bp.105.020107
Carlson, J. M., Greenberg, T., Rubin, D., & Mujica-Parodi, L. R. (2011). Feeling
anxious: Anticipatory amygdalo-insular response predicts the feeling of
anxious anticipation. Social Cognitive and Affective Neuroscience, 6(1), 74–
81. doi:10.1093/scan/nsq017
Chambless, D. L., Tran, G. Q., & Glass, C. R. (1997). Predictors of response to
cognitive-behavioral group therapy for social phobia. Journal of Anxiety
Disorders, 11(3), 221–40. http://doi.org/http://dx.doi.org/10.1016/S0887-
6185(97)00008-X
Page 52
45
Clark, D. M., & Wells, A. (1995). A cognitive model of social phobia. In R. G.
Heimberg, M. R. Liebowitz, D. A. Hope, F. R. Schneier, R. G. Heimberg, M.
R. Liebowitz, ... F. R. Schneier (Eds.) , Social phobia: Diagnosis, assessment,
and treatment (pp. 69-93). New York, NY, US: Guilford Press.
Constantino, M. J., Arnkoff, D. B., Glass, C. R., Ametrano, R. M., & Smith, J. Z.
(2011). Expectations. Journal of Clinical Psychology, 67(2), 184–192.
http://doi.org/10.1002/jclp.20754
Cooney, R. E., Atlas, L. Y., Joormann, J., Eugene, F., & Gotlib, I. H. (2006).
Amygdala activation in the processing of neutral faces in social anxiety
disorder: Is neutral really neutral? Psychiatry Research - Neuroimaging,
148(1), 55–59. doi:10.1016/j.pscychresns.2006.05.003
Craig, A. D. B. (2011). Significance of the insula for the evolution of human
awareness of feelings from the body. Annals of the New York Academy of
Sciences, 1225(1), 72–82. doi:10.1111/j.1749-6632.2011.05990.x
Cunningham, J. L., Wernroth, L., Von Knorring, L., Berglund, L., & Ekselius, L.
(2011). Agreement between physicians’ and patients' ratings on the
Montgomery-Asberg Depression Rating Scale. Journal of Affective Disorders,
135(1-3), 148–153. doi:10.1016/j.jad.2011.07.005
Damaraju, E., Huang, Y. M., Barrett, L. F., & Pessoa, L. (2009). Affective learning
enhances activity and functional connectivity in early visual cortex.
Neuropsychologia, 47(12), 2480–2487.
doi:10.1016/j.neuropsychologia.2009.04.023
Davis, M., & Whalen, P. J. (2001). The amygdala: vigilance and emotion.
Molecular Psychiatry, 6(1), 13–34. doi:10.1038/sj.mp.4000812
de Graaf, R., Bijl, R. V., Spijker, J., Beekman, A. T. F., & Vollebergh, W. A. M.
(2003). Temporal sequencing of lifetime mood disorders in relation to
comorbid anxiety and substance use disorders - Findings from the Netherlands
Mental Health Survey and Incidence Study. Social Psychiatry and Psychiatric
Epidemiology, 38(1), 1–11. doi:10.1007/s00127-003-0597-4
Page 53
46
Demenescu, L. R., Kortekaas, R., Cremers, H. R., Renken, R. J., van Tol, M. J.,
van der Wee, N. J. a, … Aleman, a. (2013). Amygdala activation and its
functional connectivity during perception of emotional faces in social phobia
and panic disorder. Journal of Psychiatric Research, 47(8), 1024–1031.
doi:10.1016/j.jpsychires.2013.03.020
Devilly, G. J., & Borkovec, T. D. (2000). Psychometric properties of the
credibility/expectancy questionnaire. Journal of Behavior Therapy and
Experimental Psychiatry, 31(2), 73–86. doi:10.1016/S0005-7916(00)00012-4
Devinsky, O., Morrell, M. J., & Vogt, B. A. (1995). Contributions of anterior
cingulate cortex to behaviour. Brain, 118(1), 279-306.
Diekhof, E. K., Geier, K., Falkai, P., & Gruber, O. (2011). Fear is only as deep as
the mind allows. A coordinate-based meta-analysis of neuroimaging studies
on the regulation of negative affect. NeuroImage, 58(1), 275–285.
doi:10.1016/j.neuroimage.2011.05.073
Doehrmann, O., Ghosh, S. S., Polli, F. E., Reynolds, G. O., Horn, F., Keshavan,
A., … Gabrieli, J. D. (2013). Predicting Treatment Response in Social
Anxiety Disorder From Functional Magnetic Resonance Imaging. JAMA
Psychiatry, 70(1), 87-97. http://doi.org/10.1001/2013.jamapsychiatry.5
Donoso, M., Collins, A. G., & Koechlin, E. (2014). Foundations of human
reasoning in the prefrontal cortex. Science, 344(6191), 1481-1486.
Doruyter, A., Lochner, C., Jordaan, G. P., Stein, D. J., Dupont, P., & Warwick, J.
M. (2016). Resting functional connectivity in social anxiety disorder and the
effect of pharmacotherapy. Psychiatry Research: Neuroimaging, 251, 34–44.
doi:10.1016/j.pscychresns.2016.04.009
Eisenberger, N. I. (2012). The pain of social disconnection: examining the shared
neural underpinnings of physical and social pain. Nature Reviews
Neuroscience, 13(6), 421–34. doi:10.1038/nrn3231
Page 54
47
Eisenberger, N. I., Master, S. L., Inagaki, T. K., Taylor, S. E., Shirinyan, D.,
Lieberman, M. D., & Naliboff, B. D. (2011). Attachment figures activate a
safety signal-related neural region and reduce pain experience. Proceedings of
the National Academy of Sciences, 108(28), 11721–11726.
http://doi.org/10.1073/pnas.1108239108
El Alaoui, S., Hedman, E., Kaldo, V., Hesser, H., Kraepelien, M., Andersson, E.,
… Lindefors, N. (2015). Effectiveness of Internet-based cognitive–behavior
therapy for social anxiety disorder in clinical psychiatry. Journal of
Consulting and Clinical Psychology, 83(5), 902–914.
http://doi.org/10.1037/a0039198
El Alaoui, S., Ljótsson, B., Hedman, E., Kaldo, V., Andersson, E., Rück, C., …
Lindefors, N. (2015). Predictors of symptomatic change and adherence in
internet-based cognitive behaviour therapy for social anxiety disorder in
routine psychiatric care. PLoS ONE, 10(4), 1–18.
http://doi.org/10.1371/journal.pone.0124258
Eskildsen, A., Hougaard, E., & Rosenberg, N. K. (2010). Pre-treatment patient
variables as predictors of drop-out and treatment outcome in cognitive
behavioural therapy for social phobia: A systematic review. Nordic Journal of
Psychiatry, 64(2), 94–105. http://doi.org/10.3109/08039480903426929
Etkin A. (2010) Functional neuroanatomy of anxiety: a neural circuit perspective.
Current Topics in Behavioral Neuroscience, 2, 251–277.
Etkin, A., Egner, T., & Kalisch, R. (2011). Emotional processing in anterior
cingulate and medial prefrontal cortex. Trends in Cognitive Sciences, 15(2),
85–93. doi:10.1016/j.tics.2010.11.004
Etkin, A., & Wager, T. D. (2007). Functional neuroimaging of anxiety: a meta-
analysis of emotional processing in PTSD, social anxiety disorder, and
specific phobia. The American Journal of Psychiatry, 164(10), 1476–1488.
doi:10.1176/appi.ajp.2007.07030504
Evans, K. C., Wright, C. I., Wedig, M. M., Gold, A. L., Pollack, M. H., & Rauch,
S. L. (2008). A functional MRI study of amygdala responses to angry
schematic faces in social anxiety disorder. Depression and Anxiety, 25(6),
496–505. doi:10.1002/da.20347
Page 55
48
Fantino, B., & Moore, N. (2009). The self-reported Montgomery-Asberg
Depression Rating Scale is a useful evaluative tool in Major Depressive
Disorder. BMC Psychiatry, 9, 26. doi:10.1186/1471-244X-9-26
Fehm, L., Pelissolo, A., Furmark, T., & Wittchen, H. U. (2005). Size and burden of
social phobia in Europe. European Neuropsychopharmacology, 15(4), 453–
462. doi:10.1016/j.euroneuro.2005.04.002
First M. B., Gibbon M., Spitzer R. L., Williams J. B. W. (1997). Structured
clinical interview for DSM-IV Axis I Disorders (SCID-I). Washington, D.C.:
American Psychiatric Press,
Fracalanza, K., McCabe, R. E., Taylor, V. H., & Antony, M. M. (2014). The effect
of comorbid major depressive disorder or bipolar disorder on cognitive
behavioral therapy for social anxiety disorder. Journal of Affective Disorders,
162, 61–66. http://doi.org/10.1016/j.jad.2014.03.015
Freitas-Ferrari, M. C., Hallak, J. E. C., Trzesniak, C., Filho, A. S., Machado-de-
Sousa, J. P., Chagas, M. H. N., … Crippa, J. A. S. (2010). Neuroimaging in
social anxiety disorder: A systematic review of the literature. Progress in
Neuro-Psychopharmacology and Biological Psychiatry, 34(4), 565–580.
doi:10.1016/j.pnpbp.2010.02.028
Frick, A, Howner, K., Fischer, H., Kristiansson, M., & Furmark, T. (2013). Altered
fusiform connectivity during processing of fearful faces in social anxiety
disorder. Translational Psychiatry, 3(10), e312. doi:10.1038/tp.2013.85
Fu, C. H. Y., Williams, S. C. R., Cleare, A. J., Scott, J., Mitterschiffthaler, M. T.,
Walsh, N. D., … Murray, R. M. (2008). Neural Responses to Sad Facial
Expressions in Major Depression Following Cognitive Behavioral Therapy.
Biological Psychiatry, 64(6), 505–512.
http://doi.org/10.1016/j.biopsych.2008.04.033
Fullana, M. A, Harrison, B. J., Soriano-Mas, C., Vervliet, B., Cardoner, N., Àvila-
Parcet, A., & Radua, J. (2016). Neural signatures of human fear conditioning:
an updated and extended meta-analysis of fMRI studies. Molecular
Psychiatry, 21(4), 500–508. doi:10.1038/mp.2015.88
Page 56
49
Furmark, T. (2002). Social phobia: overview of community surveys. Acta
Psychiatrica Scandinavica, 105(2), 84–93. doi:10.1034/j.1600-
0447.2002.1r103.x
Furmark, T., Carlbring, P., Hedman, E., Sonnenstein, A., Clevberger, P., Bohman,
B., … Andersson, G. (2009). Guided and unguided self-help for social anxiety
disorder: Randomised controlled trial. British Journal of Psychiatry, 195(5),
440–447. http://doi.org/10.1192/bjp.bp.108.060996
Furmark, T., Holmström, A., Sparthan, E., Carlbring, P., and Andersson, G.,
(2013). Social fobi: effektiv hjälp med kognitiv beteendeterapi. (2nd., [updated
and extended] edition.) Malmö: Liber.
Furmark, T., Tillfors, M., Everz, P. O., Marteinsdottir, I., Gefvert, O., &
Fredrikson, M. (1999). Social phobia in the general population: prevalence
and sociodemographic profile. Social psychiatry and psychiatric
epidemiology, 34(8), 416-424.
Furmark, T., Tillfors, M., Marteinsdottir, I., Fischer, H., Pissiota, A., Långström,
B., & Fredrikson, M. (2002). Common changes in cerebral blood flow in
patients with social phobia treated with citalopram or cognitive-behavioral
therapy. Archives of General Psychiatry, 59(5), 425–433.
doi:10.1001/archpsyc.59.5.425
Gaebler, M., Daniels, J. K., Lamke, J.P., Fydrich, T., & Walter, H. (2014).
Behavioural and neural correlates of self-focused emotion regulation in social
anxiety disorder. Journal of Psychiatry & Neuroscience : JPN, 39(4), 249–58.
doi:10.1503/jpn.130080
Gentili, C., Gobbini, M. I., Ricciardi, E., Vanello, N., Pietrini, P., Haxby, J. V., &
Guazzelli, M. (2008). Differential modulation of neural activity throughout
the distributed neural system for face perception in patients with Social
Phobia and healthy subjects. Brain Research Bulletin, 77(5), 286–292.
doi:10.1016/j.brainresbull.2008.08.003
Page 57
50
Gentili, C., Ricciardi, E., Gobbini, M. I., Santarelli, M. F., Haxby, J. V, Pietrini, P.,
& Guazzelli, M. (2009). Beyond amygdala: Default Mode Network activity
differs between patients with social phobia and healthy controls. Brain
Research Bulletin, 79(6), 409–13.
http://doi.org/10.1016/j.brainresbull.2009.02.002
Goldin, P. R., & Gross, J. J. (2010). Effects of mindfulness-based stress reduction
(MBSR) on emotion regulation in social anxiety disorder. Emotion, 10(1), 83-
91. doi:10.1037/a0018441
Goldin, P. R., Ziv, M., Jazaieri, H., Hahn, K., Heimberg, R., & Gross, J. J. (2013).
Impact of cognitive behavioral therapy for social anxiety disorder on the
neural dynamics of cognitive reappraisal of negative self-beliefs: randomized
clinical trial. JAMA psychiatry, 70(10), 1048-1056.
Goldin, P. R., Ziv, M., Jazaieri, H., Weeks, J., Heimberg, R. G., & Gross, J. J.
(2014). Impact of cognitive-behavioral therapy for social anxiety disorder on
the neural bases of emotional reactivity to and regulation of social evaluation.
Behaviour Research and Therapy, 62, 97–106. doi:10.1016/j.brat.2014.08.005
Gren-Landell, M., Tillfors, M., Furmark, T., Bohlin, G., Andersson, G., & Svedin,
C. G. (2009). Social phobia in Swedish adolescents: Prevalence and gender
differences. Social Psychiatry and Psychiatric Epidemiology, 44(1), 1–7.
doi:10.1007/s00127-008-0400-7
Hayward, C., Killen, J. D., Kraemer, H. C., & Taylor, C. B. (1998). Linking Self-
Reported Childhood Behavioral Inhibition to Adolescent Social Phobia.
Journal of the American Academy of Child & Adolescent Psychiatry, 37(12),
1308–1316. doi:10.1097/00004583-199812000-00015
Hedman, E., Andersson, E., Ljótsson, B., Andersson, G., Andersson, E., Schalling,
M., & ... Rück, C. (2012). Clinical and genetic outcome determinants of
internet‐ and group‐based cognitive behavior therapy for social anxiety
disorder. Acta Psychiatrica Scandinavica, 126(2), 126-136.
doi:10.1111/j.1600-0447.2012.01834.x
Page 58
51
Hedman, E., Andersson, G., Ljotsson, B., Andersson, E., Rück, C., Mörtberg, E.,
& Lindefors, N. (2011). Internet-based cognitive behavior therapy vs.
cognitive behavioral group therapy for social anxiety disorder: A randomized
controlled non-inferiority trial. PLoS ONE, 6(3), e18001.
http://dx.doi.org/10.1371/journal.pone.0018001
Hedman, E., Furmark, T., Carlbring, P., Ljótsson, B., Rück, C., Lindefors, N., &
Andersson, G. (2011). A 5-year follow-up of internet-based cognitive
behavior therapy for social anxiety disorder. Journal of Medical Internet
Research, 13(2).
Hedman, E., Ljotsson, B., & Lindefors, N. (2012). Cognitive behavior therapy via
the Internet: a systematic review of applications, clinical efficacy and cost-
effectiveness. Expert Review of Pharmacoeconomics & Outcomes Research,
12(6), 745–764. http://doi.org/http://dx.doi.org/10.1586/erp.12.67
Heeger, D. J., & Ress, D. (2002). What does fMRI tell us about neuronal activity?
Nature Reviews Neuroscience, 3(2), 142-151.
Heimberg, R. G., & Becker, R. E. (2002). Cognitive-behavioral group therapy for
social phobia. Basic mechanisms and clinical strategies. New York: Guilford
Press.
Heimberg, R. G., Hofmann, S. G., Liebowitz, M. R., Schneier, F. R., Smits, J. A.
J., Stein, M. B., … Craske, M. G. (2014). Social anxiety disorder in DSM-5.
Depression and Anxiety, 31(6), 472–479. doi:10.1002/da.22231
Heimberg, R. G., Stein, M. B., Hiripi, E., & Kessler, R. C. (2000). Trends in the
prevalence of social phobia in the United States: a synthetic cohort analysis of
changes over four decades. European Psychiatry : The Journal of the
Association of European Psychiatrists, 15(1), 29–37. doi:10.1016/S0924-
9338(00)00213-3
Hermann, C., Ofer, J., & Flor, H. (2004). Covariation bias for ambiguous social
stimuli in generalized social phobia. Journal of Abnormal Psychology, 113(4),
646–653. doi:10.1037/0021-843X.113.4.646
Page 59
52
Heuer, K., Rinck, M., & Becker, E. S. (2007). Avoidance of emotional facial
expressions in social anxiety: The Approach-Avoidance Task. Behaviour
Research and Therapy, 45(12), 2990-3001. doi:10.1016/j.brat.2007.08.010
Hipp, J. F., & Siegel, M. (2015). BOLD fMRI correlation reflects frequency-
specific neuronal correlation. Current Biology, 25(10), 1368–1374.
doi:10.1016/j.cub.2015.03.049
Isenberg, N., Silbersweig, D., Engelien, a, Emmerich, S., Malavade, K., Beattie,
B., … Stern, E. (1999). Linguistic threat activates the human amygdala.
Proceedings of the National Academy of Sciences of the United States of
America, 96(18), 10456–10459.doi:10.1073/pnas.96.18.10456
Jacobson, N. S., & Truax, P. (1991). Clinical significance: A statistical approach to
defining meaningful change in psychotherapy research. Journal of Consulting
and Clinical Psychology, 59(1), 12–19. doi:10.1037/0022-006X.59.1.12
Janet, P. (1903) Les obsessions et la psychasthenie. Paris: Alcan
Jenkinson, M., Beckmann, C. F., Behrens, T. E., Woolrich, M. W., & Smith, S. M.
(2012). FSL. Neuroimage, 62(2), 782-790.
Jung, J., Kang, J., Won, E., Nam, K., Lee, M. S., Tae, W. S., & Ham, B. J. (2014).
Impact of lingual gyrus volume on antidepressant response and
neurocognitive functions in Major Depressive Disorder: A voxel-based
morphometry study. Journal of Affective Disorders, 169, 179–187.
http://doi.org/10.1016/j.jad.2014.08.018
Kastner, S., & Ungerleider, L. G. (2000). Mechanisms of visual attention in the
human cortex. Annual Review of Neuroscience, 23, 315–41.
doi:10.1146/annurev.neuro.23.1.315
Kawaguchi, A., Watanabe, N., Nakano, Y., Ogawa, S., Suzuki, M., Kondo, M., …
Akechi, T. (2013). Group cognitive behavioral therapy for patients with
generalized social anxiety disorder in Japan: outcomes at 1-year follow up and
outcome predictors. Neuropsychiatric Disease and Treatment, 9, 267–75.
http://doi.org/10.2147/NDT.S41365
Page 60
53
Kennerley, S. W., Walton, M. E., Behrens, T. E. J., Buckley, M. J., & Rushworth,
M. F. S. (2006). Optimal decision making and the anterior cingulate cortex.
Nature Neuroscience, 9(7), 940–947. doi:10.1038/nn1724
Kerns, J. G., Cohen, J. D., MacDonald, A. W., Cho, R. Y., Stenger, V. A., &
Carter, C. S. (2004). Anterior cingulate conflict monitoring and adjustments in
control. Science, 303(5660), 1023-1026. doi:10.1126/science.1089910
Kessler, R. C., Berglund, P., Demler, O., Jin, R., Merikangas, K. R., & Walters, E.
E. (2005). Lifetime prevalence and age-of-onset distributions of DSM-IV
disorders in the National Comorbidity Survey Replication. Archives of general
psychiatry, 62(6), 593-602.
Klumpp, H., Angstadt, M., Nathan, P. J., & Phan, K. L. (2010). Amygdala
reactivity to faces at varying intensities of threat in generalized social phobia:
an event-related functional MRI study. Psychiatry Research: Neuroimaging,
183(2), 167-169.
Klumpp, H., Angstadt, M., & Phan, K. L. (2012). Insula reactivity and connectivity
to anterior cingulate cortex when processing threat in generalized social
anxiety disorder. Biological Psychology, 89(1), 273–276.
doi:10.1016/j.biopsycho.2011.10.010
Klumpp, H., Fitzgerald, D. A., & Phan, K. L. (2013). Neural predictors and
mechanisms of cognitive behavioral therapy on threat processing in social
anxiety disorder. Progress in Neuro-Psychopharmacology and Biological
Psychiatry, 45, 83–91. http://doi.org/10.1016/j.pnpbp.2013.05.004
Klumpp, H., Post, D., Angstadt, M., Fitzgerald, D. A, & Phan, K. L. (2013).
Anterior cingulate cortex and insula response during indirect and direct
processing of emotional faces in generalized social anxiety disorder. Biology
of Mood & Anxiety Disorders, 3(1), 7. doi:10.1186/2045-5380-3-7
Klumpp, H., Keutmann, M. K., Fitzgerald, D. A, Shankman, S. A, & Phan, K. L.
(2014). Resting state amygdala-prefrontal connectivity predicts symptom
change after cognitive behavioral therapy in generalized social anxiety
disorder. Biology of Mood & Anxiety Disorders, 4(1), 14.
http://doi.org/10.1186/s13587-014-0014-5
Page 61
54
Koenigsberg, H. W., Fan, J., Ochsner, K. N., Liu, X., Guise, K., Pizzarello, S., ...
& New, A. (2010). Neural correlates of using distancing to regulate emotional
responses to social situations. Neuropsychologia, 48(6), 1813-1822.
Kohn, N., Eickhoff, S. B., Scheller, M., Laird, A. R., Fox, P. T., & Habel, U.
(2014). Neural network of cognitive emotion regulation - An ALE meta-
analysis and MACM analysis. NeuroImage, 87, 345–355.
doi:10.1016/j.neuroimage.2013.11.001
Kolling, N., Behrens, T. E. J., Wittmann, M. K., & Rushworth, M. F. S. (2016).
Multiple signals in anterior cingulate cortex. Current Opinion in
Neurobiology, 37, 36–43. doi:10.1016/j.conb.2015.12.007
Konnopka, A., Leichsenring, F., Leibing, E., & König, H.H. (2009). Cost-of-illness
studies and cost-effectiveness analyses in anxiety disorders: a systematic
review. Journal of Affective Disorders, 114(1), 14–31.
doi:10.1016/j.jad.2008.07.014
Koric, L., Volle, E., Seassau, M., Bernard, F. A., Mancini, J., Dubois, B., … Levy,
R. (2012). How cognitive performance-induced stress can influence right
VLPFC activation: An fMRI study in healthy subjects and in patients with
social phobia. Human Brain Mapping, 33(8), 1973–1986.
doi:10.1002/hbm.21340
Kriegeskorte, N., Simmons, W. K., Bellgowan, P. S., & Baker, C. I. (2009).
Circular analysis in systems neuroscience: the dangers of double dipping.
Nature Neuroscience, 12(5), 535-540.
Kross, E., Berman, M. G., Mischel, W., Smith, E. E., & Wager, T. D. (2011).
Social rejection shares somatosensory representations with physical pain.
Proceedings of the National Academy of Sciences of the United States of
America, 108(15), 6270–6275. doi:10.1073/pnas.1102693108
Labuschagne, I., Phan, K. L., Wood, A., Angstadt, M., Chua, P., Heinrichs, M., ...
& Nathan, P. J. (2012). Medial frontal hyperactivity to sad faces in
generalized social anxiety disorder and modulation by oxytocin. The
International Journal of Neuropsychopharmacology, 15(7), 883-896.
Page 62
55
LeDoux, J. (2003). The emotional brain, fear, and the amygdala. Cellular and
molecular neurobiology, 23(4-5), 727-738.
Legrand, D., & Ruby, P. (2009). What is self-specific? Theoretical investigation
and critical review of neuroimaging results. Psychological Review, 116(1),
252–282. doi:10.1037/a0014172
Liao, W., Xu, Q., Mantini, D., Ding, J., MacHado-De-Sousa, J. P., Hallak, J. E. C.,
… Chen, H. (2011). Altered gray matter morphometry and resting-state
functional and structural connectivity in social anxiety disorder. Brain
Research, 1388, 167–177. doi:10.1016/j.brainres.2011.03.018
Lindhiem, O., Kolko, D. J., & Cheng, Y. (2012). Predicting psychotherapy benefit:
A probabilistic and individualized approach. Behavior Therapy, 43(2), 381-
392. doi:10.1016/j.beth.2011.08.004
Lindquist, K. A., Wager, T. D., Kober, H., Bliss-Moreau, E., & Barrett, L. F.
(2012). The brain basis of emotion: A meta-analytic review. Behavioral and
Brain Sciences, 35(3), 121–143. doi:10.1017/S0140525X11000446
Lissek, S., Bradford, D. E., Alvarez, R. P., Burton, P., Espensen-Sturges, T.,
Reynolds, R. C., & Grillon, C. (2014). Neural substrates of classically
conditioned fear-generalization in humans: A parametric fMRI study. Social
Cognitive and Affective Neuroscience, 9(8), 1134–1142.
doi:10.1093/scan/nst096
Lissek, S., Levenson, J., Biggs, A. L., Johnson, L. L., Ameli, R., Pine, D. S., &
Grillon, C. (2008). Elevated fear conditioning to socially relevant
unconditioned stimuli in social anxiety disorder. American Journal of
Psychiatry, 165(1), 124–132. doi:10.1176/appi.ajp.2007.06091513
Lloyd, D. (2002). Functional MRI and the study of human consciousness. Journal
of Cognitive Neuroscience, 14(6), 818–831.
doi:10.1162/089892902760191027
Loerinc, A. G., Meuret, A. E., Twohig, M. P., Rosenfield, D., Bluett, E. J., &
Craske, M. G. (2015). Response rates for CBT for anxiety disorders: Need for
standardized criteria. Clinical Psychology Review, 42, 72–82.
http://doi.org/10.1016/j.cpr.2015.08.004
Page 63
56
Logothetis, N. K., Pauls, J., Augath, M., Trinath, T., & Oeltermann, a. (2001).
Neurophysiological investigation of the basis of the fMRI signal. Nature,
412(6843), 150–7. doi:10.1038/35084005
Lueken, U., Straube, B., Konrad, C., Strohle, A., Wittmann, A., Pfleiderer, B., …
Kircher, T. (2013). Neural substrates of treatment response to cognitive-
behavioral therapy in panic disorder with agoraphobia. American Journal of
Psychiatry, 170(11), 1345–1355.
http://doi.org/http://dx.doi.org/10.1176/appi.ajp.2013.12111484
Manning, W. H., & DuBois, P. H. (1962). Correlational methods in research on
human learning. Perceptual and Motor Skills, 15, 287–321.
Månsson, K. N. T., Carlbring, P., Frick, A., Engman, J., Olsson, C. J., Bodlund, O.,
… Andersson, G. (2013). Altered neural correlates of affective processing
after internet-delivered cognitive behavior therapy for social anxiety disorder.
Psychiatry Research - Neuroimaging, 214(3), 229–237.
doi:10.1016/j.pscychresns.2013.08.012
Månsson, K. N., Frick, A., Boraxbekk, C. J., Marquand, A. F., Williams, S. C. R.,
Carlbring, P., ... & Furmark, T. (2015). Predicting long-term outcome of
Internet-delivered cognitive behavior therapy for social anxiety disorder using
fMRI and support vector machine learning. Translational psychiatry, 5(3),
e530.
Månsson, K. N. T., Salami, A., Frick, A., Carlbring, P., Andersson, G., Furmark,
T., & Boraxbekk, C.-J. (2016). Neuroplasticity in response to cognitive
behavior therapy for social anxiety disorder. Translational Psychiatry, 6(2),
e727. doi:10.1038/tp.2015.218
Mayo-Wilson, E., Dias, S., Mavranezouli, I., Kew, K., Clark, D. M., Ades, A. E.,
& Pilling, S. (2014). Psychological and pharmacological interventions for
social anxiety disorder in adults: A systematic review and network meta-
analysis. The Lancet Psychiatry, 1(5), 368–376. http://doi.org/10.1016/S2215-
0366(14)70329-3
Milad, M. R., & Quirk, G. J. (2012). Fear Extinction as a Model for Translational
Neuroscience: Ten Years of Progress. Annual Review of Psychology, 63(1),
129–151. doi:10.1146/annurev.psych.121208.131631
Page 64
57
Moffitt, T. E., Caspi, A., Taylor, A., Kokaua, J., Milne, B. J., Polanczyk, G., &
Poulton, R. (2010). How common are common mental disorders? Evidence
that lifetime prevalence rates are doubled by prospective versus retrospective
ascertainment. Psychological Medicine, 40(6), 899-909.
doi:10.1017/S0033291709991036
Mogg, K., Bradley, B., Miles, F., & Dixon, R. (2004). Time course of attentional
bias for threat scenes: Testing the vigilance‐avoidance hypothesis. Cognition
& Emotion, 18(5), 689-700. doi:10.1080/02699930341000158
Moitra, E., Beard, C., Weisberg, R. B., & Keller, M. B. (2011). Occupational
impairment and social anxiety disorder in a sample of primary care
patients.Journal of affective disorders, 130(1), 209-212.
Montgomery, S. A., & Asberg, M. (1979). A new depression scale designed to be
sensitive to change. The British journal of psychiatry, 134(4), 382-389.
Moratti, S., & Keil, A. (2009). Not what you expect: Experience but not
expectancy predicts conditioned responses in human visual and supplementary
cortex. Cerebral Cortex, 19(12), 2803–2809. doi:10.1093/cercor/bhp052
Morris, J. S., Buchel, C., & Dolan, R. J. (2001). Parallel neural responses in
amygdala subregions and sensory cortex during implicit fear conditioning.
Neuroimage, 13(6), 1044-1052.
Motzkin, J. C., Philippi, C. L., Wolf, R. C., Baskaya, M. K., & Koenigs, M.
(2015). Ventromedial prefrontal cortex is critical for the regulation of
amygdala activity in humans. Biological Psychiatry, 77(3), 276–284.
doi:10.1016/j.biopsych.2014.02.014
Mörtberg, E., & Andersson, G. (2014). Predictors of response to individual and
group cognitive behaviour therapy of social phobia. Psychology and
Psychotherapy: Theory, Research and Practice, 87(1), 32-43.
Müller, V. I., Habel, U., Derntl, B., Schneider, F., Zilles, K., Turetsky, B. I., &
Eickhoff, S. B. (2010). Incongruence effects in crossmodal emotional
integration. NeuroImage, 54(3), 2257–2266.
doi:10.1016/j.neuroimage.2010.10.047
Page 65
58
Nieuwenhuys, R. (2011). The insular cortex: a review. Progress in brain research,
195, 123-163.
Nitschke, J. B., Sarinopoulos, I., Oathes, D. J., Johnstone, T., Whalen, P. J.,
Davidson, R. J., & Kalin, N. H. (2009). Anticipatory activation in the
amygdala and anterior cingulate in generalized anxiety disorder and prediction
of treatment response. The American Journal Of Psychiatry, 166(3), 302-310.
doi:10.1176/appi.ajp.2008.07101682
Nordgreen, T., Havik, O. E., Öst, L. G., Furmark, T., Carlbring, P., & Andersson,
G. (2012). Outcome predictors in guided and unguided self-help for social
anxiety disorder. Behaviour Research and Therapy, 50(1), 13–21.
http://doi.org/10.1016/j.brat.2011.10.009
Northoff, G., Heinzel, A., De Greck, M., Bermpohl, F., Dobrowolny, H., &
Panksepp, J. (2006). Self-referential processing in our brain—a meta-analysis
of imaging studies on the self. Neuroimage, 31(1), 440-457.
Northoff, G., Qin, P., & Feinberg, T. E. (2011). Brain imaging of the self–
conceptual, anatomical and methodological issues. Consciousness and
cognition, 20(1), 52-63.
Ochsner, K. N., Silvers, J. A., & Buhle, J. T. (2012). Functional imaging studies of
emotion regulation: a synthetic review and evolving model of the cognitive
control of emotion. Annals of the New York Academy of Sciences,1251(1), E1-
E24.
Ogawa, S., Lee, T. M., Kay, A. R., & Tank, D. W. (1990). Brain magnetic
resonance imaging with contrast dependent on blood oxygenation.
Proceedings of the National Academy of Sciences, 87(24), 9868-9872.
Olatunji, B. O., Cisler, J. M., & Deacon, B. J. (2010). Efficacy of cognitive
behavioral therapy for anxiety disorders: a review of meta-analytic
findings.Psychiatric Clinics of North America, 33(3), 557-577.
Olatunji, B. O., Cisler, J. M., & Tolin, D. F. (2007). Quality of life in the anxiety
disorders: A meta-analytic review. Clinical Psychology Review, 27(5), 572–
581. doi:10.1016/j.cpr.2007.01.015
Page 66
59
Pannekoek, J. N., Veer, I. M., van Tol, M. J., van der Werff, S. J., Demenescu, L.
R., Aleman, A., ... & van der Wee, N. J. (2013). Resting-state functional
connectivity abnormalities in limbic and salience networks in social anxiety
disorder without comorbidity. European neuropsychopharmacology, 23(3),
186-195.
Pessoa, L. (2008). On the relationship between emotion and cognition. Nature
Reviews Neuroscience, 9(2), 148-158. doi:10.1038/nrn2317
Peterson, A., Thome, J., Frewen, P., & Lanius, R. A. (2014). Resting-State
Neuroimaging Studies: A New Way of Identifying Differences and
Similarities Among the Anxiety Disorders?. Canadian Journal Of Psychiatry,
59(6), 294-300.
Phan, K.L., Coccaro, E.F., Angstadt, M., Kreger, K.J., Mayberg, H.S., Liberzon, I.,
Stein, M. B. (2013). Corticolimbic brain reactivity to social signals of threat
before and after sertraline treatment in generalized social phobia. Biological
Psychiatry, 141(4), 329–336. doi:10.1016/j.surg.2006.10.010.Use
Phelps, E. A., & LeDoux, J. E. (2005). Contributions of the amygdala to emotion
processing: From animal models to human behavior. Neuron, 48(2), 175–187.
doi:10.1016/j.neuron.2005.09.025
Prater, K. E., Hosanagar, A., Klumpp, H., Angstadt, M., & Phan, K. L. (2013).
Aberrant amygdala-frontal cortex connectivity during perception of fearful
faces and at rest in generalized social anxiety disorder. Depression and
Anxiety, 30(3), 234–241. doi:10.1002/da.22014
Price, M., & Anderson, P. L. (2012). Outcome expectancy as a predictor of
treatment response in cognitive behavioral therapy for public speaking fears
within social anxiety disorder. Psychotherapy, 49(2), 173–179.
http://doi.org/10.1037/a0024734
Pujol, J., Giménez, M., Ortiz, H., Soriano-Mas, C., López-Solà, M., Farré, M., …
Martín-Santos, R. (2013). Neural response to the observable self in social
anxiety disorder. Psychological Medicine, 43(4), 721–31.
doi:10.1017/S0033291712001857
Page 67
60
Qin, P., & Northoff, G. (2011). How is our self related to midline regions and the
default-mode network? NeuroImage, 57(3), 1221–1233.
doi:10.1016/j.neuroimage.2011.05.028
Qiu, C., Liao, W., Ding, J., Feng, Y., Zhu, C., Nie, X., … Gong, Q. (2011).
Regional homogeneity changes in social anxiety disorder: A resting-state
fMRI study. Psychiatry Research - Neuroimaging, 194(1), 47–53.
doi:10.1016/j.pscychresns.2011.01.010
Quadflieg, S., Mohr, A., Mentzel, H. J., Miltner, W. H. R., & Straube, T. (2008).
Modulation of the neural network involved in the processing of anger
prosody: The role of task-relevance and social phobia. Biological Psychology,
78(2), 129–137. doi:10.1016/j.biopsycho.2008.01.014
Raichle, M. E. (2015). The brain's default mode network. Annual review of
neuroscience, 38, 433-447.
Rescorla, R. A, & Wagner, AR. (1972). A theory of Pavlovian conditioning:
variations in the effectiveness of reinforcement and nonreinforcement. In AH.
Black & W.F. Prokasy (eds.), Classical conditioning II: current research and
theory (pp. 64-99) New York: Appleton-Century-Crofts.
Rodebaugh, T. L., Holaway, R. M., & Heimberg, R. G. (2004). The treatment of
social anxiety disorder. Clinical Psychology Review, 24(7), 883–908.
http://doi.org/10.1016/j.cpr.2004.07.007
Ronchi, R., Bello-Ruiz, J., Lukowska, M., Herbelin, B., Cabrilo, I., Schaller, K., &
Blanke, O. (2015). Right insular damage decreases heartbeat awareness and
alters cardio-visual effects on bodily self-consciousness. Neuropsychologia,
70, 11–20. doi:10.1016/j.neuropsychologia.2015.02.010
Rorden, C., & Brett, M. (2000). Stereotaxic display of brain lesions. Behavioural
Neurology, 12(4), 191.
Rosellini, A. J., Rutter, L. A., Bourgeois, M. L., Emmert-Aronson, B. O., &
Brown, T. A. (2013). The relevance of age of onset to the psychopathology of
social phobia. Journal of Psychopathology and Behavioral Assessment, 35(3),
356–365. doi:10.1007/s10862-013-9338-5
Page 68
61
Ruben, J., Schwiemann, J., Deuchert, M., Meyer, R., Krause, T., Curio, G., …
Villringer, A. (2001). Somatotopic organization of human secondary
somatosensory cortex. Cerbral Cortex, 11(5), 463–473.
Schibbye, P., Ghaderi, A., Ljótsson, B., Hedman, E., Lindefors, N., Rück, C., &
Kaldo, V. (2014). Using early change to predict outcome in cognitive
behaviour therapy: Exploring timeframe, calculation method, and differences
of disorder-specific versus general measures. PLoS ONE, 9(6).
http://doi.org/10.1371/journal.pone.0100614
Schmidt, S., Mohr, A., Miltner, W. H. R., & Straube, T. (2010). Task-dependent
neural correlates of the processing of verbal threat-related stimuli in social
phobia. Biological Psychology, 84(2), 304–312.
doi:10.1016/j.biopsycho.2010.03.005
Scholing, A., & Emmelkamp, P. M. (1999). Prediction of treatment outcome in
social phobia: a cross-validation. Behaviour Research and Therapy, 37(7),
659–70. http://doi.org/10.1016/S0005-7967(98)00175-2
Sehlmeyer, C., Schöning, S., Zwitserlood, P., Pfleiderer, B., Kircher, T., Arolt, V.,
& Konrad, C. (2009). Human fear conditioning and extinction in
neuroimaging: A systematic review. PLoS ONE, 4(6).
doi:10.1371/journal.pone.0005865
Shah, S. G., Klumpp, H., Angstadt, M., Nathan, P. J., & Phan, K. L. (2009).
Amygdala and insula response to emotional images in patients with
generalized social anxiety disorder. Journal Of Psychiatry & Neuroscience,
34(4), 296-302.
Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E.,
& ... Dunbar, G. C. (1998). The Mini-International Neuropsychiatric Interview
(M.I.N.I): The development and validation of a structured diagnostic
psychiatric interview for DSM-IV and ICD-10. Journal Of Clinical
Psychiatry, 59(Suppl 20), 22-33.
Shin, L. M., & Liberzon, I. (2010). The neurocircuitry of fear, stress, and anxiety
disorders. Neuropsychopharmacology, 35(1), 169-191.
doi:10.1038/npp.2009.83
Page 69
62
Siegle, G. J., Carter, C. S., & Thase, M. E. (2006). Use of fMRI to predict recovery
from unipolar depression with cognitive behavior therapy. American Journal
of Psychiatry, 163(4), 735–738. http://doi.org/10.1176/appi.ajp.163.4.735
Smith, R., & Lane, R. D. (2015). The neural basis of one’s own conscious and
unconscious emotional states. Neuroscience and Biobehavioral Reviews, 57,
1–29. doi:10.1016/j.neubiorev.2015.08.003
Somers, J. M., Goldner, E. M., Waraich, P., & Hsu, L. (2006). Prevalence and
incidence studies of anxiety disorders: A systematic review of the literature.
Canadian Journal of Psychiatry, 51(2), 100–113.
Sripada, C. S., Angstadt, M., Banks, S., Nathan, P. J., Liberzon, I., & Phan, K. L.
(2009). Functional neuroimaging of mentalizing during the trust game in
social anxiety disorder. Neuroreport, 20(11), 984–9.
doi:10.1097/WNR.0b013e32832d0a67
Statistics Sweden. (2015). Physical and mental health by indicator, age and sex.
Percentage and estimated numbers in thousands. Year 2008-2009 - 2014-
2014. Living Conditions Surveys (ULF/SILC). Retrieved 11 febuary, 2016
from
http://www.statistikdatabasen.scb.se/pxweb/en/ssd/START__LE__LE0101__
LE0101H/LE0101H01/?rxid=1647089b-3a76-46c2-87a7-9eb0001e26cd
Stein, M. B., & Gorman, J. M. (2001). Unmasking social anxiety disorder. Journal
Of Psychiatry & Neuroscience, 26(3), 185-189.
Stein, M. B., Simmons, A. N., Feinstein, J. S., & Paulus, M. P. (2007). Increased
amygdala and insula activation during emotion processing in anxiety-prone
subjects. American Journal of Psychiatry, 164(2), 318–327.
doi:10.1176/appi.ajp.164.2.318
Steketee, G., & Chambless, D. L. (1992). Methodological issues in prediction of
treatment outcome. Clinical Psychology Review, 12(4), 387-400.
Page 70
63
Sung, S. C., Porter, E., Robinaugh, D. J., Marks, E. H., Marques, L. M., Otto, M.
W., & ... Simon, N. M. (2012). Mood regulation and quality of life in social
anxiety disorder: An examination of generalized expectancies for negative
mood regulation. Journal Of Anxiety Disorders, 26(3), 435-441.
doi:10.1016/j.janxdis.2012.01.004
Swedish Council on Health Technology Assessment [SBU]. (2005). Behandling av
ångestsyndrom: en systematisk litteraturöversikt Vol. 1, from
http://www.sbu.se/upload/Publikationer/Content0/1/angest_ Vol_1.pdf
Tillfors, M., Furmark, T., Carlbring, P., & Andersson, G. (2015). Risk profiles for
poor treatment response to internet-delivered CBT in people with social
anxiety disorder. Journal of Anxiety Disorders, 33, 103–9.
http://doi.org/10.1016/j.janxdis.2015.05.007
van der Meer, L., Costafreda, S., Aleman, A., & David, A. S. (2010). Self-
reflection and the brain: a theoretical review and meta-analysis of
neuroimaging studies with implications for schizophrenia. Neuroscience and
Biobehavioral Reviews, 34(6), 935–946.
Vogt, B. A. (2005). Pain and emotion interactions in subregions of the cingulate
gyrus. Nature Reviews. Neuroscience, 6(7), 533–544. doi:10.1038/nrn1704
Vogt, B. A. (2016). Midcingulate Cortex: Structure, Connections, Homologies,
Functions and Diseases. Journal of Chemical Neuroanatomy, 74, 28–46.
doi:10.1016/j.jchemneu.2016.01.010
Vogt, B. A., Berger, G. R., & Derbyshire, S. W. G. (2003). Structural and
functional dichotomy of human midcingulate cortex. European Journal of
Neuroscience, 18(11), 3134–3144. doi:10.1111/j.1460-9568.2003.03034.
Wager, T. D., Lindquist, M., & Kaplan, L. (2007). Meta-analysis of functional
neuroimaging data: Current and future directions. Social Cognitive and
Affective Neuroscience, 2(2), 150–158. doi:10.1093/scan/nsm015
Weiller, E., Bisserbe, J. C., Boyer, P., Lepine, J. P., & van Lecrubier, Y. (1996).
Social phobia in general health care: an unrecognised undertreated disabling
disorder. The British Journal Of Psychiatry: The Journal Of Mental Science,
168(2), 169-174.
Page 71
64
Whitfield-Gabrieli, S., Ghosh, S. S., Nieto-Castanon, A., Saygin, Z., Doehrmann,
O., Chai, X. J., … Gabrieli, J. D. E. (2015). Brain connectomics predict
response to treatment in social anxiety disorder. Molecular Psychiatry, 21, 1–
6. http://doi.org/10.1038/mp.2015.109
Wittchen, H. U., Fuetsch, M., Sonntag, H., Müller, N., & Liebowitz, M. (1999).
Disability and quality of life in pure and comorbid social phobia - Findings
from a controlled study. European Psychiatry, 14(3), 118–131.
doi:10.1016/S0924-9338(99)80729-9
Wittchen, H. U., & Jacobi, F. (2005). Size and burden of mental disorders in
Europe - A critical review and appraisal of 27 studies. European
Neuropsychopharmacology, 15(4), 357–376.
doi:10.1016/j.euroneuro.2005.04.012
Wong, N., Sarver, D. E., & Beidel, D. C. (2012). Quality of life impairments
among adults with social phobia: The impact of subtype. Journal Of Anxiety
Disorders, 26(1), 50-57. doi:10.1016/j.janxdis.2011.08.012
Wood, J. N., & Grafman, J. (2003). Human prefrontal cortex: processing and
representational perspectives. Nature Reviews Neuroscience, 4(2), 139–147.
doi:10.1038/nrn1033
Yonkers, K. A., Dyck, I. R., & Keller, M. B. (2001). An eight-year longitudinal
comparison of clinical course and characteristics of social phobia among men
and women. Psychiatric Services, 52(5), 637–643.
doi:10.1176/appi.ps.52.5.637
Ziv, M., Goldin, P. R., Jazaieri, H., Hahn, K. S., & Gross, J. J. (2013). Is there less
to social anxiety than meets the eye? Behavioral and neural responses to three
socio-emotional tasks. Biology of Mood & Anxiety Disorders, 3, 1–14.
doi:10.1186/2045-5380-3-5
Page 72
65
Appendix A
Liebowitz skala för social ångest (LSAS-SR) Fyll i följande formulär med det mest passande svarsalternativet som listas nedan. Basera svaren på hur
du har känt dig den senaste veckan (tänk dig in i hur du skulle reagera om du inte har varit med om
situationen under veckan). Om du har fyllt i formuläret tidigare, försök vara konsekvent i hur du bedömer
varje situation. Var vänlig och svara på alla frågor.
Rädsla eller ångest Undvikande
0 = Ingen 0 = Aldrig (0%)
1 = Mild 1 = Ibland (1–33% av tiden)
2 = Måttlig 2 = Ofta (33–67% av tiden)
3 = Stark 3 = Vanligtvis (67–100% av tiden)
Gäller för följande situationer:
Rädsla eller
ångest (0-3)
Undvik
ande
(0-3)
1. Tala i telefon på allmän plats
2. Delta i små grupper
3. Äta på offentliga platser
4. Dricka tillsammans med andra på offentlig plats
5. Tala med människor i maktposition
6. Tala eller uppträda inför publik
7. Gå på fest
8. Utföra ett arbete medan någon ser på
9. Skriva medan någon ser på
10. Ringa upp någon som du inte känner särskilt väl
11. Tala med personer du inte känner väl
12. Träffa främmande människor
13. Urinera på en offentlig toalett
14. Komma in i ett rum där andra redan sitter ned
15. Vara i centrum för andras uppmärksamhet
16. Ta ordet på ett möte
17. Göra ett prov
18. Uttrycka oenighet eller ogillande med en person som du inte känner
särskilt väl
19. Se personer i ögonen som du inte känner särskilt väl
20. Avge en rapport till en grupp
21. Göra försök till sexuell kontakt
22. Lämna tillbaka (klaga på) köpta varor i en affär
23. Ordna en bjudning hemma
24. Stå emot en påstridig försäljare
Page 73
66
Appendix B
Montgomery Åsberg Depression Rating Scale (MADRS-S)
Namn:________________________________________________ Datum:_______________
Avsikten med de följande 9 frågorna är att ge en detaljerad bild av Ditt nuvarande sinnestillstånd.
Vi vill alltså att Du ska försöka gradera hur Du mått under de senaste tre dygnen.
Frågorna innehåller en rad olika påståenden om hur man kan må i olika avseenden. Påståendena
uttrycker olika grader av obehag, från frånvaro av obehag till maximalt uttalat obehag.
Markera det alternativ Du tycker bäst stämmer in på hur Du mått de senaste tre dagarna. Tänk inte allt
för länge, utan försök arbeta snabbt.
Fråga Svar
1. Sinnestämning
Här ber vi dig beskriva din sinnesstämning, om du känner dig ledsen, tungsint eller dyster till
mods. Tänk efter hur du har känt dig de senaste tre dagarna, om du har skiftat i humöret eller
om det varit i stort sätt detsamma hela tiden, och försök särskilt komma ihåg om du känt dig
lättare till sinnet och det hänt något positivt.
0 Jag kan känna mig glad eller ledsen, alltefter omständigheterna.
1
2 Jag känner mig nedstämd för det mesta, men ibland kan det kännas lättare.
3
4 Jag känner mig genomgående nedstämd och dyster. Jag kan inte glädja mig åt sådant som
vanligen skulle göra mig glad.
5
6 Jag är totalt nedstämd och så olycklig att jag inte kan tänka mig värre.
2. Orokänslor
Här ber vi dig markera i vilken utsträckning du haft känslor av inre spänning, olust och ångest
eller odefinierad rädsla under de senaste tre dagarna. Tänk särskilt på hur intensiva känslorna
varit, om de kommit och gått eller funnits nästan hela tiden.
0 Jag känner mig mestadels lugn.
1
2 Ibland har jag obehagliga känslor av inre oro.
3
4 Jag har ofta en känsla av inre oro, som ibland kan bli mycket stark, och som jag måste
anstränga mig för att bemästra.
5
6 Jag har fruktansvärda, långvariga eller outhärdliga ångestkänslor.
3. Sömn
Här ber vi dig beskriva hur du sover. Tänk efter hur länge du sovit och hur god sömnen varit
under de senaste tre nätterna. Bedömningen skall avse hur du faktiskt sovit, oavsett om du tagit
sömnmedicin eller ej. Om du sover mer än vanligt, markera det första alternativet.
0
Jag sover lugnt och bra och tillräckligt länge för mina behov. Jag har inga särskilda svårigheter
att somna.
Page 74
67
1
2 Jag har vissa sömnsvårigheter. Ibland har jag svårt att somna eller sover ytligare eller
oroligare än vanligt.
3
4 Jag sover minst två timmar mindre per natt än normalt. Jag vaknar ofta under natten, även
om jag inte blir störd.
5
6 Jag sover mycket dåligt, inte mer än 2–3 timmar per natt.
4. Matlust
Här ber vi dig ta ställning till hur din aptit är, och tänka efter om den på något sätt skilt sig från
vad som är normalt för dig. Om du skulle ha bättre aptit än normalt, markera då första
alternativet.
0 Min aptit är som den brukar vara.
1
2 Min aptit är sämre än vanligt.
3
4 Min aptit har nästan helt försvunnit. Maten smakar inte och jag måste tvinga mig att äta.
5
6 Jag vill inte ha någon mat. Om jag ska få någonting i mig, måste jag övertalas att äta
5. Koncentrationsförmåga
Här ber vi dig ta ställning till din förmåga att hålla tankarna samlade och koncentrera dig på
olika aktiviteter. Tänk igenom hur du fungerar vid olika sysslor som kräver olika grad av
koncentrationsförmåga, till exempel läsning av komplicerad text, lätt tidningstext och TV-
tittande.
0
Jag har inga koncentrationssvårigheter.
1
2 Jag har tillfälligt svårt att hålla tankarna samlade på sådant som normalt skulle fånga min
uppmärksamhet (t.ex. läsning eller tv-tittande).
3
4 Jag har påtagligt svårt att koncentrera mig på sådant som normalt inte kräver någon
ansträngning från min sida (t.ex. läsning eller samtal med andra människor).
5
6 Jag kan överhuvudtaget inte koncentrera mig på någonting.
6. Initiativförmåga
Här ber vi dig att försöka värdera din handlingskraft. Frågan gäller om du har lätt eller svårt för
att komma igång med sådant som du tycker du bör göra, och i vilken utsträckning du måste
övervinna ett inre motstånd när du ska ta itu med något.
0
Jag har inga svårigheter med att ta itu med nya uppgifter.
1
2 När jag ska ta itu med något, tar det emot på ett sätt som inte är normalt för mig.
3
4 Det krävs en stor ansträngning för mig att ens komma igång med enkla uppgifter som jag
vanligtvis utför mer eller mindre rutinmässigt.
5
Page 75
68
6 Jag kan inte förmå mig att ta itu med de enklaste vardagsbestyr.
7. Känslomässigt engagemang
Här ber vi dig ta ställning till hur du upplever ditt intresse för omvärlden och för andra
människor, och för sådana aktiviteter som brukar bereda dig nöje och glädje.
0 Jag är intresserad av omvärlden och engagerar mig i den, och det bereder mig både nöje och
glädje.
1
2 Jag känner mindre starkt för sådant som brukar engagera mig. Jag har svårare än vanligt att
bli glad eller svårare att bli arg när det är befogat.
3
4 Jag kan inte känna något intresse för omvärlden, inte ens för vänner och bekanta.
5
6 Jag har slutat uppleva några känslor. Jag känner mig smärtsamt likgiltig även för mina
närmaste.
8. Pessimism
Frågan gäller hur du ser på din egen framtid och hur du uppfattar ditt eget värde. Tänk efter i
vilken utsträckning du ger dig självförebråelser, om du plågas av skuldkänslor, och om du
oroat dig oftare än vanligt för till exempel din ekonomi eller din hälsa.
0 Jag ser på framtiden med tillförsikt. Jag är på det hela taget ganska nöjd med mig själv.
1
2 Ibland klandrar jag mig själv och tycker att jag är mindre värd än andra.
3
4 Jag grubblar ofta över mina misslyckanden och känner mig mindervärdig eller dålig, även
om andra tycker annorlunda.
5
6 Jag ser allting i svart och kan inte se någon ljusning. Det känns som om jag var en alltigenom
dålig människa, och som om jag aldrig skulle kunna få någon förlåtelse för det hemska jag
gjort.
9. Livslust
Frågan gäller din livslust, och om du känt livsleda. Har du tankar på självmord, och i så fall, i
vilken utsträckning upplever du detta som en verklig utväg?
0 Jag har normal aptit på livet.
1
2 Livet känns inte särskilt meningsfullt, men jag önskar ändå inte att jag vore död.
3
4 Jag tycker ofta det vore bättre att vara död, och trots att jag egentligen inte önskar det, kan
självmord ibland kännas som en möjlig utväg.
5
6 Jag är egentligen övertygad om att min enda utväg är att dö, och jag tänker mycket på hur jag
bäst ska gå tillväga för att ta mitt eget liv.
Page 76
69
Appendix C
CEQ Frågor om behandlingen för social rädsla
Dina svar kommer inte att visas för din behandlare. Markera med ett kryss under den siffra som motsvarar
din uppfattning bäst.
1. Hur logisk tycker du att den här typen av behandling verkar?
0=Inte alls logisk 10=Mycket logisk
0 1 2 3 4 5 6 7 8 9 10
Internet-baserad
kognitiv beteendeterapi
2. Hur säker är du på att den här metoden kommer vara framgångsrik i behandlingen av din sociala
rädsla?
0=Inte alls säker 10=Mycket säker
0 1 2 3 4 5 6 7 8 9 10
Internet-baserad
kognitiv beteendeterapi
3. Med vilken grad av tillit skulle du rekommendera den här behandlingsmetoden till en vän med samma
typ av problem som du har?
0=Inte alls tillitsfull 10=Mycket tillitsfull
0 1 2 3 4 5 6 7 8 9 10
Internet-baserad
kognitiv beteendeterapi
4. Hur framgångsrik tror du att denna behandling skulle vara i behandling av andra rädslor av olika
slag?
0=Inte alls framgångsrik 10=Mycket framgångsrik
0 1 2 3 4 5 6 7 8 9 10
Internet-baserad
kognitiv beteendeterapi
5. Hur förbättrad förväntar du dig bli av den här behandlingen?
0=Ingen förbättring alls 10=Helt bra/symptomfri
0 1 2 3 4 5 6 7 8 9 10
Internet-baserad
kognitiv beteendeterapi
Page 77
70
Appendix D
Experimental task stimuli
27 statements directed in second person, and to a significant other in third person signified
with the placeholder “x” below. The Original statements are in Swedish seen here and
translated to English below in Appendix E.
Self-referential criticism Other-referential criticism
Du är en idiot X är en idiot
Du är tråkig X är tråkig
Du gör bort dig X gör bort sig
Du ser konstig ut X ser konstig ut
Alla skrattar åt dig Alla skrattar åt X
Du är en nolla X är en nolla
Du är pinsam X är pinsam
Alla hatar dig Alla hatar X
Du är misslyckand X är misslyckand
Alla tittar på dig Alla tittar på X
Ingen vill ha dig Ingen vill ha X
Du gör alltid fel X gör alltid fel
Du rodnar X rodnar
Du är värdelös X är värdelös
Du är korkad X är korkad
Du är dum X är dum
Du klarar det aldrig X klarar det aldrig
Du är ingenting X är ingenting
Du är svag X är svag
Du beter dig konstigt X beter sig konstigt
Du gör aldrig rätt X gör aldrig något rätt
Du är klumpig X är klumpig
Du är ful X är ful
Du är hatad X är hatad
Ingen respekterar dig Ingen respekterar X
Du är en förlorare X är en förlorare
Du är nervös X är nervös
Page 78
71
Appendix E
Experimental task stimuli
27 statements directed in second person, and to a significant other in third person signified
with the placeholder “X” below. The Original statements are in Swedish, see Appendix D,
translated to English below.
Self-referential criticism Other-referential criticism
You are an idiot X is an idiot
You are boring X is boring
You are making a fool out of yourself
X
is making a fool out of
themselves
You look weird X looks weird
Everyone is laughing at you Everyone is laughing at X
You are a nobody X is a nobody
You are embarrassing X is embarrassing
Everyone hates you Everyone hates X
You are a failure X is a failure
Everyone is watching you Everyone is watching X
Nobody wants you Nobody wants X
You always screw up X always screws up
You are blushing X is blushing
You are worthless X is worthless
You are stupid X is stupid
You are dumb X is dumb
You will never make it X will never make it
You are nothing X is nothing
You are weak X is weak
You are acting strange X is acting strange
You never do anything right
X
never does anything
right
You are clumsy X is clumsy
You are ugly X is ugly
You are hated X is hated
No one respects you No one respects X
You are a looser X is a loser
You are nervous X is nervous
Page 79
72
Appendix F
Whole-brain univariate voxel-wise test to Other > Self and vise versa
Table F1
Whole-brain univariate voxel-wise test to Other > Self and vise versa.
MNI-
coordinates
Volume
Contrast Region x y z (mm3) Z
Other > Self
L Anterior middle temporal gyrusa -60 -12 -14 2192 5.22
L Dorsal posterior cingulate cortexa -4 -54 28 5904 4.99
R Middle occipital gyrusa 6 -90 -10 2168 4.6
R Angular gyrus 46 -58 22 1296 4.34
L Angular gyrusa -42 -58 22 2616 4.27
L Ventromedial prefrontal cortex -4 58 -10 160 3.84
L Supplementary motor area -8 20 56 432 3.79
L Middle occipital gyrus -12 -100 -2 408 3.65
L Posterior middle temporal gyrus -58 -34 -2 352 3.47
L Ventromedial prefrontal cortex -6 66 26 296 3.4
L Dorsomedial prefrontal cortex -6 48 40 264 3.31
R Supplementary motor area 18 32 46 64 3.22
L Subgenual cingulate cortex -2 8 -8 16 3.17
R Middle occipital gyrus 18 -80 -16 16 3.14
L Ventromedial prefrontal cortex -6 64 0 24 3.12
Self > Other
R Superior supramarginal gyrusa 48 -34 50 2232 4.36
L Supramarginal areaa -64 -42 34 7960 4.23
R Inferior supramarginal gyrus 66 -32 22 1784 3.83
R Dorsal posterior cingulate cortex 12 -32 36 480 3.79
R Postcentral gyrus 60 -20 34 352 3.66
R Occipitotemporal area 50 -58 -4 336 3.54
R Posterior middle temporal gyrus 54 -44 2 152 3.47
L Hippocampus -30 -40 8 40 3.35
R Inferior supplementary motor area 52 0 6 24 3.27
L Occipitotemporal area -46 -68 0 56 3.25
L Precentral gyrus -44 -6 14 32 3.24
R Dorsal posterior cingulate cortex 28 -54 12 40 3.24
R Intraparietal sulcus 38 -50 10 48 3.24
L Premotor cortex -60 4 22 16 3.2
R Dorsal posterior cingulate cortex 16 -18 36 16 3.17
Note. All listed clusters significant at p < 0.001 (uncorrected), MNI, Montreal Neurological
Institute; Z, Z-score. a pFWE < 0.05
Page 80
73
Appendix G
Fear circuitry activation in Other > Fixation Cross and Self > Fixation Cross
contrasts
Figure G1 Figure G2
Other > Fixation Cross. Showing only activity
in the amygdala, hippocampus, and insula
bilaterally. All shown clusters significant at p <
0.001 (uncorrected). Sagittal view, x = -20,
Coronal view, y = -32, Axial view, z = -4
Self > Fixation Cross. Showing only activity in
the amygdala, hippocampus, and insula
bilaterally. All shown clusters significant at p <
0.001 (uncorrected). Sagittal view, x = -20,
Coronal view, y = -32, Axial view, z = -4
Page 81
74
Appendix H
Correlations between predictor variables
Table H1
Correlations between predictor variables
Measure 1 2 3 4 5 6
1. Age –
2. Gender -0.27 –
3. MADRS-S -0.01 0.23 –
4. CEQ -0.09 0.08 0.16 –
5. L pMCC 0.00 0.00 -0.13 0.27 –
6. L LG 0.20 -0.09 -0.03 0.16 0.42** –
Note. ** Correlations significant at the 0.01 level (2-tailed). MADRS-S, Montgomery-Åsberg
Depression Rating Scale; CEQ, Credibility/Expectancy Questionnaire; pMCC, Posterior Mid
Cingulate Cortex; LG, Lingual Gyrus.