Social Anxiety
Contents lists available at ScienceDirect
Journal of Affective Disorders
journal homepage: www.elsevier.com/locate/jad
Research paper
The role of expressive suppression and cognitive reappraisal in cognitive behavioral therapy for social anxiety disorder: A study of self-report, subjective, and electrocortical measures Yogev Kivity⁎,1, Lior Cohen, Michal Weiss, Jonathan Elizur, Jonathan D. Huppert Department of Psychology, The Hebrew University of Jerusalem, Jerusalem, Israel
A R T I C L E I N F O
Keywords: Social Anxiety Emotion Regulation Cognitive Reappraisal Expressive Suppression Cognitive Behavioral Therapy
A B S T R A C T
Background: Contemporary models of cognitive behavioral therapy (CBT) for social anxiety disorder (SAD) emphasize emotion dysregulation as a core impairment whose reduction may play a causal role in psy- chotherapy. The current study examined changes in use of emotion regulation strategies as possible mechanisms of change in CBT for SAD. Specifically, we examined changes in expressive suppression and cognitive reappraisal during CBT and whether these changes predict treatment outcome. Methods: Patients (n = 34; 13 females; Mean age = 28.36 (6.97)) were allocated to 16-20 sessions of CBT. An electrocortical measure of emotion regulation and a clinician-rated measure of SAD were administered monthly. Self-report measures of emotion regulation and social anxiety were administered weekly. Multilevel models were used to examine changes in emotion regulation during treatment and cross-lagged associations between emotion regulation and anxiety. Results: CBT led to decreased suppression frequency, increased reappraisal self-efficacy, and decreased un- pleasantness for SAD-related pictures (ps < .05). At post-treatment, patients were equivalent to healthy controls in terms of suppression frequency and subjective reactivity to SAD-related stimuli. Gains were maintained at 3- months follow-up. Decreases in suppression frequency and electrocortical reactivity to SAD-related pictures predicted lower subsequent anxiety but not the other way around (ps < .05). Lower anxiety predicted greater subsequent increases in reappraisal self-efficacy. Limitations: The lack of a control group precludes conclusions regarding mechanisms specificity. Conclusions: Decreased frequency of suppression is a potential mechanism of change in CBT for SAD.
1. Introduction
Recent models of anxiety, including social anxiety disorder (SAD), emphasize impairments in emotion regulation (Hofmann, Sawyer, Fang, & Asnaani, 2012; Morrison & Heimberg, 2013). Two regulation strategies, cognitive reappraisal and expressive suppression, may be particularly relevant for SAD (Morrison & Heimberg, 2013). In the process model of emotion regulation (Gross, 2015), cognitive re- appraisal is generally considered an adaptive strategy that involves cognitive change to regulate one's emotion – for example, attempts to reinterpret emotional stimuli in less threatening ways (Gross, 2015). On the other hand, expressive suppression is an attempt to inhibit one's expression of emotions and is generally considered maladaptive (Gross, 2015).
In Heimberg's updated model (Morrison & Heimberg, 2013),
emotion dysregulation in SAD includes avoidance of anxiety (e.g. avoidance or escape from stressful situations) and expressive suppres- sion due to believing that expressing emotions will lead to rejection or to excessive focus on oneself. The model further proposes that in- dividuals with SAD are less effective in implementing reappraisal. Ac- cordingly, decreased suppression and increased effective use of re- appraisal are hypothesized to lead to symptom reduction, for example, by outward shifting of attention and by reducing exaggerated prob- ability and cost of rejection.
Recently, studies have examined suppression and reappraisal in SAD (reviewed in Dryman & Heimberg, 2018). Cross-sectional and daily diary studies typically focus on the frequency of use of a strategy and self-efficacy (perceived success in implementation). Individuals with SAD report an over-reliance on suppression and lower frequency and self-efficacy of reappraisal (e.g., Farmer & Kashdan, 2012;
https://doi.org/10.1016/j.jad.2020.10.021 Received 10 May 2020; Received in revised form 16 August 2020; Accepted 11 October 2020
⁎ Corresponding author: Yogev Kivity, Department of Psychology, Bar Ilan University, Ramat Gan 5290002, Israel, Telephone: +972-3-5318715 E-mail address: [email protected] (Y. Kivity).
1 Yogev Kivity is now in the Department of Psychology, Bar Ilan University, Israel.
Journal of Affective Disorders 279 (2021) 334–342
Available online 14 October 2020 0165-0327/ © 2020 Elsevier B.V. All rights reserved.
T
Gaebler, Daniels, Lamke, Fydrich, & Walter, 2014). Impairments in frequency seem to be larger than impairments in self-efficacy in sup- pression while the opposite is true for reappraisal (Kivity & Huppert, 2018, 2019), thus supporting the Heimberg model.
In addition, studies have also utilized subjective ratings during lab tasks of emotion regulation to study reappraisal and suppression abil- ities in SAD. These studies typically present SAD-related stimuli to participants, such as pictures of rejecting faces (Goldin et al., 2009a), pictures portraying scenes of rejection and embarrassment (Kivity & Huppert, 2016, 2018, 2019) and idiographic statements of negative self-beliefs (Goldin et al., 2009b) while asking participants to change the way they interpret these stimuli in order to reduce the distress they evoke in them. However, compared to questionnaires and daily diary measures, these lab studies have shown intact emotion regulation abilities in SAD compared to controls, even under social stress (e.g., Gaebler et al., 2014; Kivity & Huppert, 2016, 2018, 2019). Thus, self- reported impairments are not reflected in lab performance. Among other possibilities, this discrepancy may suggest a difficulty im- plementing strategies in daily life despite an intact ability to implement them upon instruction in controlled circumstances, low ecological va- lidity of lab-based measures, or a bias in self-reports that does not exist in lab-based measures. Examining the role that each of these aspects (lab-based performance, self-reported frequency and self-reported self- efficacy) plays in treatments for SAD may shed light on their relative importance.
Several techniques of cognitive behavioral therapy (CBT) for SAD seem relevant for improving emotion regulation. Psychoeducation and exposures likely decrease suppression, as patients learn that hiding their anxiety is futile and likely to backfire. Outward shifting of at- tention (focusing on the task at hand instead of on how one is per- ceived) presumably decreases suppression by decreasing patients’ pre- occupation with their overt signs of anxiety. Furthermore, cognitive restructuring can potentially increase the use of reappraisal by chan- ging biased catastrophic cognitions. Finally, psychoeducation and in- vivo exposure challenge biased cognitions and are expected to promote reappraisal too.
Studies have shown that self-reported reappraisal (frequency and self-efficacy) increases in CBT (Goldin et al., 2014a; Goldin, Morrison, Jazaieri, Heimberg, & Gross, 2017; Kocovski, Fleming, Hawley, Huta, & Antony, 2013; Moscovitch et al., 2012). However, findings regarding self-reported suppression are inconclusive, with one study reporting a decrease in frequency (Goldin et al., 2014a) and another reporting no change (Moscovitch et al., 2012). Less is known regarding lab-based measures: one study found improvements in reappraisal of negative social evaluations and negative self-beliefs during CBT (Goldin et al., 2013, 2014b).
Importantly, the best test of the importance of emotion regulation as a treatment target for SAD is to examine its contribution to symptom improvement (Nock, 2007). Changes in reappraisal and suppression that predict treatment outcome would provide further support to Heimberg's model. Several studies found that increases in self-reported frequency and self-efficacy of reappraisal predicted subsequent symptom reduction (Goldin et al., 2017; Kocovski, Fleming, Hawley, Ho, & Antony, 2015; Moscovitch et al., 2012), although another study found that only self-efficacy (but not frequency) of reappraisal pre- dicted subsequent outcome (Goldin et al., 2014a). Decreases in self- reported suppression frequency predicted contemporaneous, but not subsequent, symptoms reduction in one study (Goldin et al., 2014a) and did not predict outcome at all in another (Moscovitch et al., 2012). The only examination of lab-based reappraisal (Goldin et al., 2014b) found that greater changes in fMRI measures of reappraisal predicted greater symptom change during CBT for SAD, although subjective task per- formance did not. Thus, the self-report findings suggest that reappraisal increases during CBT for SAD and may be driving symptom change, with more consistent findings regarding self-efficacy than frequency. Additional studies are needed regarding lab-based emotion regulation.
The present study examines changes in suppression and reappraisal during CBT for SAD and their role in treatment outcome using data from a previously completed study (Huppert, Kivity, Cohen, Strauss, Elizur & Weiss, 2018). We collected weekly self-reports of the frequency and self-efficacy of reappraisal and suppression, and monthly lab-based measures of reappraisal and suppression in response to SAD-related pictures. Given past findings, our primary focus was self-reported self- efficacy of reappraisal and frequency of suppression. Frequency of re- appraisal and self-efficacy of suppression were of secondary interest.
For the lab-based task, we focused on unpleasantness and event- related potentials (ERP) during reappraisal and suppression of SAD- related pictures. We focused on the late positive potential (LPP), an ERP with a posterior midline distribution and an onset of 300 ms post-sti- mulus (Proudfit, Dunning, Foti, & Weinberg, 2013). Larger LPP reflect sustained attention to stimuli and elaborative engagement in order to regulate the emotion (Proudfit et al., 2013). The LPP is sensitive to emotional intensity and to emotion regulation, with higher amplitudes for highly arousing stimuli that are reduced following instructions to regulate, including reappraisal and suppression, even within several seconds from the presentation of the stimuli (Proudfit et al., 2013). The LPP shows less habituation over repeated exposure to stimuli compared to other psychophysiological measures (Proudfit et al., 2013), which allows for repetition of stimuli and attribution of changes in LPPs to the same stimuli as due to regulation. The LPP has been utilized in SAD (Kinney, Burkhouse, & Klump, 2019; Kivity & Huppert, 2018, 2019; Yuan et al., 2014) but we are unaware of studies examining it during CBT for SAD.
We examined these hypotheses: 1) CBT will result in significant improvements in suppression and reappraisal, including reduced fre- quency of suppression and increased self-efficacy, frequency, and suc- cessful lab implementation of reappraisal. 2) Improvements in sup- pression and reappraisal will play a potentially causal role in CBT: improvements will predict subsequent improvements in anxiety and not vice versa. 3) We examined whether patients reached an equivalent level of emotion regulation to healthy controls (HCs) at post-treatment, and whether gains were maintained at 3-months follow-up without an a-priori hypothesis.
2. Method
2.1. Participants
Data were drawn from the CBT arm of a study of treatments for SAD (Huppert et al., 2018)2. Patients were recruited via advertisements and referrals. Participants were 34 patients who met DSM-IV-TR (American Psychiatric Association, 2000) criteria for SAD and 40 HCs with no history of psychiatric disorders, matched to patients on sex, age and education. One patient decided not to enter treatment and 5 HCs were removed because they did not have a continued low social anxiety score between screening and participation. The final sample included 33 patients (13 females, Age: 18-53, M = 28.36, SD = 6.97) and 35 HCs (15 females, Age: 19-45, M = 28.49, SD = 6.28). Participants were Hebrew speaking and family status was: single: CBT = 59%, HC = 61%; in a relationship: CBT = 41%, HC = 35%; divorced: CBT = 0%, HC = 4%. Education levels were: high school: CBT = 15%, HC = 29%; post-high school: CBT = 21%, HC = 9%; undergraduate degree/student: CBT = 42%, HC = 27%; graduate degree/student:
2 The original study also included participants receiving a computerized treatment for SAD called Attention Bias Modification. This treatment was of shorter duration compared to CBT, of a smaller sample size and only included three measurements of lab-based emotion regulation. In addition, group as- signment was random only for a subset of the CBT patients. Due to these rea- sons, we decided not to include data from this treatment in the current study, which a priori was designed to examine the role of ER in CBT.
Y. Kivity, et al. Journal of Affective Disorders 279 (2021) 334–342
335
CBT = 21%, HC = 26%. Groups did not differ on demographics (ps > .05). Ten SAD participants (29.41%) had one comorbid disorder and two (5.88%) had more than one. The most common comorbid disorders were depression (n = 9; 26.47%) or other anxiety disorders (n = 4; 11.76%).
2.2. Measures
Liebowitz Social Anxiety Scale (LSAS; Liebowitz, 1987). A 24- item interviewer-rated measure of fear and avoidance of social inter- actions and social performance. The Hebrew version (Levin, Marom, Gur, Wechter, & Hermesh, 2002) was administered by trained clinical psychology doctoral students, blind to hypotheses. Internal consistency in all assessments was α = .90 – .96; interrater reliability for 15 ran- domly chosen interviews was r = .94.
Social Phobia Inventory (SPIN; Connor et al., 2000). A valid and reliable 17-item self-report measure of social anxiety symptoms, translated and back translated to Hebrew for previous studies. Internal consistency in all assessments was: α = .78 – .95.
Emotion Regulation Questionnaire – Self-Efficacy and Frequency (ERQ; Gross & John, 2003). We used the reliable and valid Hebrew version (Carthy, Horesh, Apter, Edge, & Gross, 2010). Fol- lowing Goldin et al. (2009b), we measured both the frequency (ERQ-F) and self-efficacy (ERQ-SE) in social situations instead of frequency only (internal consistency of all subscales: α = .73 – .97). Items tap into reappraisal (e.g., “When I want to feel less negative emotion, I change the way I'm thinking about the situation”) and suppression (e.g., “I control my emotions by not expressing them”) which participants en- dorse using a 1 ("Seldom"/"Ineffectively") to 7 ("Often"/"Effectively") scale.
Emotion Regulation Task. Full details are provided in supple- mental material, section 1. We used a task that was developed by Hajcak and Nieuwenhuis (2006) who presented emotionally-salient pictures to participants and instructed them to either passively view the picture or to reappraise the emotion it evokes in them while ERP ac- tivity was recorded and unpleasantness ratings were collected. Hajcak and Nieuwenhuis found that the amplitude of the LPP and the level of subjective unpleasantness were decreased during reappraisal compared to passive viewing. Thus, the task is validated and suitable for studying the effects of emotion regulation on electrocortical activity and sub- jective unpleasantness. In the current study, we adapted the task to measure suppression in addition to reappraisal and used SAD-related pictures rather than general pictures (Kivity & Huppert, 2018, 2019). In selecting the stimuli for the task, we chose to focus on shame, embar- rassment and rejection because these experiences are central in SAD (Goldin et al., 2009b; Morrison & Heimberg, 2013; Moscovitch, 2009)3. Twenty trials of each condition were included: viewing of SAD-related pictures, viewing of neutral pictures, reappraisal of SAD-related pic- tures and suppression of SAD-related pictures.
When viewing SAD-related pictures, participants imagined them- selves as the character that is the focus of shame, rejection, and em- barrassment. When viewing neutral pictures, participants responded naturally. When reappraising, participants first imagined themselves as the character and then changed the way they think of the picture to decrease their unpleasantness (e.g., "This guy is not laughing at me, but at someone else"). When suppressing, participants first imagined themselves as the character and then concealed any expression of emotions. To enhance the effect of the suppression manipulation, a web
camera was placed above the computer screen and participants were told that a member of the research team would review the recordings. Participants were instructed to avoid any expression of their emotions such that it would be impossible to tell whether they were viewing neutral pictures or concealing their emotions4.
After each trial, participants rated their unpleasantness on a Self- Assessment Manikin (SAM; Lang, Bradley, & Cuthbert, 2008) scale (1 through 9; 5 being neutral; transformed such that higher ratings express greater unpleasantness). We focused on unpleasantness ratings in order to complement the LPP data (which is mostly correlated with arousal) and arrive at a more complete picture of the participants’ emotional experience that takes into account the two basic dimensions of emo- tions – valence and arousal. After providing unpleasantness ratings, participants were asked to indicate the instructions they followed during that trial. In the reappraisal condition, participants were also asked to record the new interpretation they came up with for the pic- ture (results not reported here).
Ratings were averaged for each condition and a regulation score (view – regulate; calculated on the transformed scores) was calculated to capture the amount of reduction in unpleasantness. Higher scores indicate larger regulation effects.
SAD-related pictures were collected from the internet5, normed and shown to evoke moderate shame, embarrassment, rejection and un- pleasantness (Kivity & Huppert, 2018, 2019). These depicted situations of shame, rejection, and/or embarrassment such as scenes of people pointed and laughed at, anxious people during a public speech, and facial expressions of contempt. Neutral pictures were taken from the International Affective Picture System (IAPS) database (Lang et al., 2008).
Psychophysiological Recording, Data Reduction, and Analysis. Full details are provided in section 2 of the supplement. ERPs were constructed by averaging trials in each condition (view, suppression, reappraisal and view neutral). Following Moser, Hartwig, Moran, Jendrusina, & Kross (2014), the LPP was quantified as the average voltage in 5 parietal electrodes (CPz, P1, Pz, P2, POz) in the entire segment (400-2000 ms). A regulation change score (view – regulate) was calculated to capture the amount of reduction in the LPP. Higher scores indicate larger regulation effects. Trials in which participants failed to use the instructed strategy were excluded (4.57% on average, no group differences). Studies have shown that the LPP can be reliably measured with as little as 8 trials and that it varies little beyond 12 trials (Moran, Jendrusina, & Moser, 2013). Assessments with fewer than 12 valid trials in each condition were removed from analyses (5% across groups, no group differences).
2.3. Treatment and therapists
Individual CBT was delivered for up to 20 sessions using a manual by Roth-Ledley, Foa, & Huppert (2006), based on Clark's (2005) CBT for SAD. Components such as building an idiographic model, outward shifting of attention, safety behaviors experiment, video feedback, be- havioral experiments and exposures, and optional use of imaginal ex- posure, assertiveness training, or social skills training are included. Therapists were four clinical psychology doctoral students with 2-4 years of CBT experience. Videorecordings of sessions were used in group supervision by the last author.
3 Shame, embarrassment and rejection are likely separate, but related, ex- periences. Similar to previous studies (e.g., Goldin et al., 2009), when designing and validating the task (Kivity & Huppert, 2018, 2019) we were not able to examine these experiences separately due to a small number of stimuli that purely fall into one of these categories. It remains for future studies to examine these experiences separately.
4 It should be noted that although the view condition is not entirely a passive one (as it includes perspective taking), it is still possible that it requires less cognitive effort than the reappraisal and suppression conditions. However, studies have shown that cognitive effort alone does not explain the down-reg- ulatory effects of reappraisal (Foti & Hajack, 2008).
5 See a sample picture at https://tinyurl.com/ShameArousingPicture.
Y. Kivity, et al. Journal of Affective Disorders 279 (2021) 334–342
336
2.4. Procedure
The institutional review board approved the study. After providing informed consent, participants were evaluated by trained independent evaluators (blind to hypotheses) using the Mini-International Neuropsychiatric Interview (Sheehan et al., 1998) and the LSAS. Par- ticipants completed a baseline assessment and entered treatment. Pa- tients completed the ERQ before and after each session (post-session ratings were of secondary interest and are only reported in Supple- mental Material, Section 4). Patients also completed in-lab assessments at pre-treatment, every four sessions during treatment, at post-treat- ment and at 3-months follow-up which included the LSAS, SPIN, ERQ, and the lab task. Thus, each patient had up to seven assessments (pre- treatment, sessions 4, 8, 12 and 16, post-treatment, and follow-up). HCs only completed a single assessment and were not followed long- itudinally. Thus, HCs were included only in analyses of equivalency.
2.5. Data Analyses
We used intent-to-treat linear multi-level models (assessments at level 1 repeated within patients at level 2) implemented in R package 'nlme' (Pinheiro, Bates, DebRoy, Sarkar, & R Core Team, 2016). In- cluding therapists as a third level showed negligible and non-significant effects (ICCs: Med = .00, range: 0 – 0.049) and therefore this level was removed. We used restricted maximum likelihood estimation, a first- order autoregressive level 1 covariance structure and random intercepts and slopes at level 2. For H1, linear rates of change were examined by including session/assessment number as a level 1 predictor (centered at pre-treatment). Intercepts represent estimated levels of the dependent variable at pre-treatment and slopes represent estimated changes in the dependent variable between two assessment points (one/four sessions, depending on the measure). To examine changes from post-treatment to follow-up we fitted a piece-wise model that examines changes during treatment and from post-treatment to follow-up separately. This was done by adding the follow-up data to the abovementioned model and adding a dummy coded variable that captures post-treatment to follow- up changes (coded “1” for follow-up assessment and “0” for all other assessment). The fixed effect of the dummy variable expresses the amount and significance of the change from post-treatment to follow- up.
H2 was examined by modeling within-patient variation in the pre- dictor following recommended procedures (Wang & Maxwell, 2015). Monthly scores of the predictor (patient mean-centered) served as within-patient scores in a cross-lagged (1-month) model. Within-subject effects represent cross-lagged associations between the predictor and the outcome. Per Wang and Maxwell (2015) we did not control for linear time effects as we wished to model and explain these very effects. Following Falkenström, Finkel, Sandell, Rubel, and Holmqvist (2017), we did not include the lagged dependent variable as a predictor because it introduces a dependency between the dependent variable and the error, thus violating assumptions. However, the first auto-regressive residual structure partly accounts for the effects of prior on current levels of the outcome. For consistency, we only analyzed monthly scores of the SPIN and ERQ. A cross-lagged association was interpreted as significant only if effects were significant for clinician-rated and self- reported anxiety.
For H3, comparisons were conducted using clinical equivalency procedures (Kendall, Marrs-Garcia, Nath, & Sheldrick, 1999) through t tests examining non-inferiority (i.e., < 1 SD difference) of post-treat- ment scores compared to HCs. A significant effect in noninferiority tests suggests that patients are non-inferior to HCs.
Effect sizes were calculated as semi-partial r (rs; Jaeger, Edwards, Das, & Sen, 2016; Nakagawa & Schielzeth, 2013) using package 'r2glmm' in R (Jaeger & R Core Team, 2016). These represent the un- ique contribution above and beyond the contribution of other pre- dictors in the model and are presented in absolute values.
3. Results
3.1. Changes in regulation (H1) and equivalency to HCs (H3)
3.1.1. Change in self-reports Descriptive statistics for all study variables are presented in sup-
plemental material, section 3. Changes are shown in Fig. 1. Suppression. Consistent with hypotheses, the frequency of sup-
pression decreased during treatment (t467 = -3.98, p < .01, rs = .22 [.14, .30]) and did not change from post-treatment to follow-up (b = -.10, t490 = -.62, p = .53, rs = .01 [.00, .10]). In contrast to hypotheses, self-efficacy of suppression decreased during treatment (t467 = -2.85, p < .01, rs = .14 [.05, .22]) and did not change from post-treatment to follow-up (b = -.09, t490 = -.64, p = .52, rs = .01 [.00, .10]).
Reappraisal. Consistent with hypotheses, self-efficacy of reappraisal increased during treatment (t467 = 3.67, p < .01, rs = .17 [.08, .25]) and did not change from post-treatment to follow-up (b = -.02, t490 = -.10, p = .92, rs = .00 [.00, .10]). In contrast to hypotheses, no changes in frequency of reappraisal were observed (t467 = .04, p = .97, rs = .00 [.00, .10]) nor did they change from post-treatment to follow- up (b = .08, t490 = .33, p = .74, rs = .01 [.00, .10]).
3.1.2. Change in lab-based measures Changes are shown in Fig. 2 and Fig. 3. View SAD-related pictures. As hypothesized, unpleasantness ratings
decreased during treatment (t115 = 4.49, p < .01, rs = .28 [.13, .42]) and did not change from post-treatment to follow-up (b = .08, t138 = .70, p = .49, rs = .03 [.00, .18]). The LPP while viewing SAD- related pictures did not change significantly during treatment (b = -.46, t103 = -1.80, p = .07, rs = .14 [.01, .30]) or from post-treatment to
Fig. 1. Change in frequency and self-efficacy of reappraisal (top panel) and suppression (bottom panel) during Cognitive Behavioral Therapy (CBT). Error bars represent estimated standard errors. Only data from sessions 1-16 are presented because only 5 patients received more than 16 sessions. b = esti- mated weekly change in emotion regulation. ** p < .01
Y. Kivity, et al. Journal of Affective Disorders 279 (2021) 334–342
337
follow-up (b = 1.73, t126 = 1.11, p = .27, rs = .05 [.00, .20]). View neutral pictures. Supporting our hypotheses, we found no
change in unpleasantness ratings during CBT or from post-treatment to follow-up (during: t115 = -.43, p = .67, rs = .04 [.00, .20]; post- treatment to follow-up: b = -.05, t138 = -.26, p = .79, rs = .02 [.00, .17]) and in the LPP (during: b = -.29, t103 = -1.04, p = .30, rs = .09 [.00, .25]; post-treatment to follow-up: b = 1.86, t126 = 1.09, p = .28, rs = .05 [.00, .20]) while viewing neutral pictures.
Suppression. Examining suppression-related reductions in un- pleasantness (compared to viewing pictures) we found that reductions were significantly different from zero at pre-treatment (b = .25, t115 = 3.84, p < .01), indicating that suppression was effective in down-regulating negativity. Consistent with hypotheses, regulation scores did not change during treatment (t115 = -1.77, p = .08, rs = .16 [.02, .31]) or from post-treatment to follow-up (b = .03, t138 = .29, p = .78, rs = .02 [.00, .18]). Reductions in LPP via suppression were not different from zero at pre-treatment (b = .69, t103 = 1.00, p = .32), suggesting that suppression was not effective in down regulating the LPP. Also consistent with hypotheses, regulation scores did not change during treatment (b = -.10, t103 = -.36, p = .72, rs = .03 [.00, .20]) and from post-treatment to follow-up (b = -.68, t126 = -.50, p = .61, rs = .03 [.00, .19]).
Reappraisal. Reappraisal-related reductions in unpleasantness were significantly different from zero at pre-treatment (b = 1.41, t115 = 10.33, p < .01) indicating that reappraisal was indeed effective in down-regulating negativity. However, in contrast to hypotheses, we also found significant decreases in regulation scores during CBT (t115 = - 3.98, p < .01, rs = .26 [.11, .41] and no change from post-treatment to follow-up (b = -.16, t138 = -1.25, p = .21, rs = .06 [.00, .21]). Additional analyses showed that raw unpleasantness ratings during reappraisal did not change (b = .02, t115 = .60, p = .55, rs = .03 [.00, .20]) which suggests that the decrease in regulation scores of re- appraisal was driven by a decrease in unpleasantness while simply viewing SAD-related pictures and not by a decrease in unpleasantness ratings during reappraisal.
Fig. 2. Mean unpleasantness ratings while viewing neutral pictures and while viewing (top panel), suppressing, and reappraising (bottom panel) SAD-related pictures during Cognitive Behavioral Therapy. Higher scores indicate greater unpleasantness. Error bars represent estimated standard errors. Only data from sessions 1-16 are presented because only 5 patients received more than 16 sessions. b = estimated change in unpleasantness every four session. ** p < .01
Fig. 3. Picture-locked Event Related Potentials in the emotion regulation task pooled at parietal sites (CPz, P1, Pz, P2, POz), shown separately at pre- (top panel) and post- (bottom panel) cognitive behavioral therapy. Green, blue, red, and black lines represent amplitudes in the view neutral, reappraisal, suppression and view conditions, respectively.
Y. Kivity, et al. Journal of Affective Disorders 279 (2021) 334–342
338
When examining reappraisal-related reductions in the LPP, we found that reductions were not different from zero at pre-treatment (b = .52, t103 = .71, p = .48), indicating that reappraisal was not effective in down-regulating the LPP. Also, inconsistent with hy- potheses, regulation scores did not change during treatment (b = -.06, t103 = -.20, p = .84, rs = .02 [.00, .19]) and from post-treatment to follow-up (b = -1.66, t126 = -1.21, p = .23, rs = .10 [.01, .25]).
3.1.3. Equivalency to HCs Equivalency tests were only carried out for measures in which pa-
tients showed impairments compared to HCs at baseline: self-report measures of emotion regulation and subjective unpleasantness while viewing SAD-related pictures (Kivity & Huppert, 2019). In terms of self- reported emotion regulation, at post-treatment, patients reached equivalency to HCs in terms of frequency and self-efficacy of suppres- sion (p < .01) but continued to show lower frequency and self-efficacy of reappraisal (ps > .05). In addition, patients reached equivalency to HCs in terms of unpleasantness while viewing SAD-related pictures (p < .02).
3.2. Cross-lagged associations between emotion regulation and anxiety (Hypothesis 2)
A summary of estimated cross-lagged fixed-effects are presented in Table 1.
3.2.1. Self-reported emotion regulation Frequency of suppression. Consistent with hypotheses, lower sup-
pression frequency predicted lower subsequent clinician-rated (t81 = 2.11, p = .04, rs = .11 [.01, .29]) and self-reported anxiety (t86 = 2.41, p = .02, rs = .18 [.02, .35]). Thus, lower suppression frequency predicted lower anxiety four sessions later. Furthermore, neither clinician-rated (t81 = 1.46, p = .15, rs = .08 [.00, .26]) nor self- reported anxiety (t86 = 1.68, p = .10, rs = .08 [.00, .26]) predicted subsequent suppression frequency. These results suggest a unidirec- tional association whereby suppression frequency predicts subsequent anxiety and not the other way around.
Frequency of reappraisal. In contrast to hypotheses, no significant cross-lagged associations were found between reappraisal frequency and anxiety (reappraisal→clinician-rated anxiety: t81 = -.39, p = .70, rs = .02 [.00, .22]; reappraisal→self-reported anxiety: t86 = -.45, p = .65, rs = .02 [.00, .21]; clinician-rated anxiety → reappraisal: t81 = -.91, p = .37, rs = .07 [.00, .25]; self-reported anxiety → re- appraisal: t86 = -.55, p = .58, rs = .04 [.00, .22]).
Self-efficacy of suppression. In partial support of hypotheses,
greater suppression self-efficacy predicted lower subsequent self-re- ported (t86 = 2.25, p = .03, rs = .17 [.02, .34]) but not clinician-rated anxiety (t81 = 1.90, p = .06, rs = .09 [.00, .27]). In addition, neither clinician-rated (t81 = .97, p = .34, rs = .04 [.00, .23]) nor self-reported anxiety (t86 = 1.82, p = .07, rs = .07 [.00, .25]) predicted subsequent suppression self-efficacy.
Self-efficacy of reappraisal. In partial support of hypotheses, greater reappraisal self-efficacy predicted lower subsequent self-re- ported (t86 = -2.09, p = .04, rs = .12 [.01, .30]) but not clinician-rated anxiety (t81 = -1.45, p = .15, rs = .06 [.00, .25]). Furthermore, in contrast to hypotheses, lower clinician-rated (t81 = 2.11, p = .04, rs = .11 [.01, .29]) and self-reported anxiety (t86 = -2.77, p < .01, rs = .11 [.01, .29]) predicted greater subsequent reappraisal self-effi- cacy. These results are more supportive of a unidirectional relationship such that anxiety predicts subsequent reappraisal self-efficacy and not vice-versa.6
3.2. 2. Lab-based emotion regulation Unpleasantness ratings and electrocortical responses while viewing
SAD-related pictures. Lower unpleasantness ratings while viewing SAD-related pictures predicted subsequent self-reported (t85 = -2.50, p = .01, rs = .18 [.02, .35]), but not clinician-rated, anxiety (t80 = - 1.46, p = .15, rs = .09 [.00, .27]). However, lower clinician-rated (t80 = -2.39, p = .02, rs = .14 [.01, .32]) and lower self-reported an- xiety (t85 = -2.46, p = .02, rs = .14 [.01, .31]) predicted lower sub- sequent unpleasantness ratings.
For electrocortical responses, smaller LPPs while viewing SAD-re- lated pictures predicted lower subsequent clinician-rated (t70 = 2.30, p = .02, rs = .12 [.01, .31]) and self-reported anxiety (t75 = 2.73, p < .01, rs = .17 [.01, .35]). In addition, anxiety did not predict subsequent LPPs while viewing SAD-related pictures (clinician-rated anxiety: b = -.02, t69 = -.36, p = .72, rs = .03 [.00, .24]; self-reported: t74 = -.72, p = .47, rs = .08 [.00, .27]). This suggests a unidirectional association whereby lower electrocortical reactivity to SAD-related stimuli predicts lower anxiety four sessions later and not the other way around.
Table 1 Estimated multi-level fixed effects and 95% confidence intervals for cross-lagged associations between emotion regulation and social anxiety.
Emotion Regulation → Anxiety Anxiety → Emotion Regulation Emotion regulation variable Clinician-rated anxiety Self-reported anxiety Clinician-rated anxiety Self-reported anxiety
ERQ Suppression frequency 2.92 [0.16, 5.67]* 2.34 [0.41, 4.28]* 0.01 [0.00, 0.02] 0.01 [0.00, 0.03] Reappraisal self-efficacy -2.55 [-6.05, 0.96] -2.46 [-4.80, -0.12]* -0.01 [-0.02, 0.00]* -0.02 [-0.03, 0.00]** Suppression self-efficacy 2.72 [-0.12, 5.57] 2.49 [0.29, 4.69]* 0.00 [-0.01, 0.01] 0.01 [0.00, 0.03] Reappraisal frequency -0.62 [-3.80, 2.56] -0.44 [-2.38, 1.50] -0.01 [-0.02, 0.01] -0.01 [-0.03, 0.01] Lab-based (ratings) View -4.22 [-9.96, 1.53] -4.41 [-7.92, -0.90]* -0.01 [-0.02, 0.00]* -0.01 [-0.02, 0.00]* Suppression -6.74 [-16.82, 3.35] 0.60 [-5.41, 6.61] 0.00 [0.00, 0.01] 0.00 [-0.01, 0.01] Reappraisal -0.30 [-5.77, 5.18] 1.99 [-1.00, 4.98] 0.01 [0.00, 0.01] 0.01 [0.00, 0.02] Lab-based (LPP) View 0.70 [0.09, 1.30]* 0.49 [0.13, 0.85]** -0.02 [-0.11, 0.08] -0.06 [-0.21, 0.10] Suppression 0.28 [-0.25, 0.81] 0.00 [-0.32, 0.31] -0.04 [-0.13, 0.05] -0.03 [-0.15, 0.09] Reappraisal -0.03 [-0.47, 0.42] 0.12 [-0.14, 0.38] 0.04 [-0.05, 0.13] -0.04 [-0.18, 0.10]
Notes. ERQ = Emotion regulation questionnaire; LPP = Late positive potential. Clinician-rated and self-reported anxiety was measured using the Liebowitz Social Anxiety Scale and the Social Phobia Inventory, respectively. Lab-based reappraisal and suppression represent regulation change scores (view – regulate) with higher scores indicating larger regulation effects. */** significantly different from zero at p < .05/.01.
6 We also examined whether our findings are sensitive to the time lag between the measurement of emotion regulation and anxiety (one vs. four weeks) by refitting cross-lagged models of self-reported emotion regulation and self-re- ported anxiety with all weekly data. Lab-based emotion regulation and clin- ician-rated anxiety were only measured every four weeks and were not included in these analyses. We found that lower anxiety predicted lower subsequent suppression frequency and greater subsequent reappraisal self-efficacy, and not the other way around. The findings are reported in full in Supplemental Material, Section 5.
Y. Kivity, et al. Journal of Affective Disorders 279 (2021) 334–342
339
Lab-based reappraisal and suppression. Reappraisal- and suppres- sion-related reductions in unpleasantness and electrocortical responses did not predict and were not predicted from clinician-rated and self- reported anxiety (ts = -1.33 – 1.88; ps = .06 – .98; rss = .00 – .12).
4. Discussion
We utilized a multi-method design to examine the role of emotion regulation in CBT for SAD. CBT led to decreased self-reported frequency and self-efficacy of suppression, increased self-reported self-efficacy of reappraisal, decreased unpleasantness while viewing SAD-related pic- tures, and smaller reappraisal-related reductions in unpleasantness while viewing SAD-related pictures. Gains were maintained at 3- months follow-up. Lower suppression frequency and smaller electro- cortical reactivity predicted lower subsequent anxiety, but not vice versa. In contrast, self-efficacy of reappraisal and subjective reactivity to SAD-related pictures were predicted by previous levels of anxiety, but not vice versa.
4.1. Changes in emotion-regulation during treatment
Self-report findings are consistent with the literature regarding in- creases in reappraisal (Goldin et al., 2014a, 2017; Kocovski et al., 2013; Moscovitch et al., 2012), and partly consistent regarding decreases in suppression (Goldin et al., 2014a). Moscovitch et al. (2012) found no changes in suppression, which could be due to a number of reasons (e.g., different CBT protocols, insufficient power). Our study is novel in showing that patients reached normal levels of suppression, but not reappraisal. Thus, CBT seems to eliminate the initial overreliance on suppression among individuals with SAD – an impairment that has been shown to be large (e.g., g = 0.96 in Kivity & Huppert, 2019). Im- plementation of reappraisal seems to take more time to develop, pos- sibly due to its complexity.
In terms of lab-based emotion regulation, we found decreased sub- jective reactivity to SAD-related pictures. Goldin et al. (2013) found similar reductions in reactivity that were specific to CBT compared with waitlist controls, and that only CBT patients reported an increase in automatic use of reappraisal during view. These findings, along with the finding that decreased subjective reactivity predicted lower sub- sequent self-reported anxiety suggest that the reduction is unlikely to be due to practice effects (i.e., habituation due to the mere repeated ex- posure to the stimuli). Rather, it is possible that as treatment progresses, patients' automatic regulation increases even when passively viewing. This is consistent with our findings that the decrease in unpleasantness during reappraisal was mostly driven by reduction in unpleasantness while viewing pictures, and not by changes in unpleasantness in the reappraisal condition. It is noteworthy that floor effects could not be ruled out because unpleasantness was low at pre-treatment. However, even at post-treatment, unpleasantness while viewing SAD-related pictures did not reach the level of neutral pictures (see supplemental material, section 3, for descriptive statistics) making floor effect a less likely explanation. Finally, the LPP during lab-based regulation did not change, which is consistent with other studies reporting that the LPP is not larger among participants with social anxiety (Kivity & Huppert, 2018. 2019; Yuan et al., 2014, but see Kinney et al., 2019 for an exception).
4.2. The association between emotion-regulation and anxiety
Lower frequency of suppression predicted lower subsequent anxiety, but not the other way around. This supports a unidirectional association such that as patients make less attempts to hide their anxiety, their anxiety level decreases. This is consistent with the emphasis in CBT on reducing attempts to hide anxiety which is conveyed in several ways throughout treatment, such as in the psychoeducation module, outward shifting of attention (instead of on one's overt signs of anxiety) and in
vivo exposures (patients are exposed to anxiety provoking situations while dropping safety behaviors that are intended to hide their an- xiety). Together with the finding that patients reached a normal level of suppression frequency, our findings point to a unique role of suppres- sion in CBT for SAD.
Surprisingly, suppression did not predict subsequent anxiety in previous studies (Goldin et al., 2014a; Moscovitch et al., 2012). This divergence in findings could be due to the specific CBT protocol used in the current study that uniquely emphasizes outward shifting of atten- tion (Clark, 2005; Roth-Ledley et al., 2006). Our findings are consistent with a recent study showing that suppression, and not reappraisal, predicts subsequent anxiety in CBT for panic disorder (Strauss, Kivity, & Huppert, 2019). In addition, in our sensitivity analyses, anxiety pre- dicted suppression in a one-week lag, but not vice versa. Thus, the exact time lag that determines the association between suppression and an- xiety needs to be further understood (c.f., Cole & Maxwell, 2003) and more studies examining the role of suppression in various CBT protocols are needed.
Self-reported reappraisal did not reliably predict subsequent anxiety but rather was predicted by it. This is in contrast to previous studies showing that reappraisal predicted subsequent anxiety (Goldin et al., 2014a, 2017; Kocovski et al., 2015; Moscovitch et al., 2012). Together with our findings regarding suppression, it seems like the association between habitual emotion regulation and anxiety in CBT is dynamic, includes bidirectional influences between regulation and anxiety (c.f. Hofmann, Curtiss, & Hayes, 2020) and depends on the specific CBT protocol being used.
Lower electrocortical reactivity to SAD-related stimuli predicted lower subsequent anxiety, but not vice versa. This unidirectional as- sociation was unique to electrocortical reactivity whereas subjective reactivity seemed to be predicted by anxiety and not vice versa. Thus, while reductions in LPPs during treatment were not significant, they did predict outcome. This may suggest that some CBT patients become less reactive to SAD-related stimuli and that decreased reactivity is related to subsequent symptom reduction. The mechanism underlying this re- duction remains to be examined. For example, the decreased reactivity could be related to use of automatic reappraisal (that is not captured by the lab-based task) or to improvements in other regulation strategies that are known to be impaired in SAD such as rumination (Kashdan & Roberts, 2007) and avoidance (Heur, Rinck, & Becker, 2007).
Other than the abovementioned findings, lab-based regulation did not improve and was unrelated to outcome. This is consistent with the lack of impairments in social anxiety compared to controls in lab-based tasks (Gaebler et al., 2014; Goldin et al., 2009b; Kinney et al., 2019; Kivity & Huppert, 2018, 2019; Yuan et al., 2014; but see Ziv, Goldin, Jazaieri, Hahn, & Gross, 2013 for a significant finding in one out of three tasks). Along with daily diaries studies showing impaired emotion regulation (e.g., Kivity & Huppert, 2016), our findings suggest that individuals with SAD are able to regulate upon instruction, but have difficulties implementing their abilities in daily life. However, one study found specific CBT-related decreases in emotion intensity ratings and modulation of fMRI activity during reappraisal of negative social evaluations (Goldin et al., 2013, 2014b). These differences could either stem from the different tasks used, the measure of brain activity (ERPs vs. fMRI), or the materials used (standardized pictures vs. idiographic self-referential social evaluations).
Taken together, our findings suggest that CBT eliminates the over- reliance on suppression and that this reduction in suppression fre- quency plays an important role in CBT for SAD whereas reappraisal does not. This is only partly consistent with the Heimberg model (Morrison & Heimberg, 2013) which attributes a role to both suppres- sion and reappraisal in CBT for SAD. Thus, as more findings accrue, the model could be modified to put a larger emphasis on suppression re- lative to reappraisal. Furthermore, models should explicitly articulate the ways in which CBT is hypothesized to reduce suppression (e.g., outward shifting of attention, dropping safety behaviors). Future
Y. Kivity, et al. Journal of Affective Disorders 279 (2021) 334–342
340
studies could empirically examine these propositions and potentially improve existing protocols. Finally, at least one other route for anxiety reduction exist in CBT that occurs through diminished reactivity to SAD-related stimuli. This remains to be further understood in future studies.
5. Limitations
Despite the multiple assessment points, our sample was moderate in size. Relatedly, unbiased crossed-lagged estimates an accurate time lag between cause and outcome (Cole & Maxwell, 2003). Such information is available for emotion regulation-social anxiety associations, and we mostly focused on four-weeks lags. Additional studies with intensive measurements are needed to determine the most accurate time lags. Finally, without a treatment control group, it is difficult to rule out that the observed changes were not due to the mere passage of time. However, the finding that changes in emotion regulation were only observed during treatment but not from post-treatment to follow-up reduce the likelihood of changes due to the passage of time only. Still, the lack of a control group makes it impossible to determine the spe- cificity of the findings to CBT compared to other treatments.
6. Conclusions
CBT for SAD reduced emotion dysregulation as revealed by self- reports and subjective emotional reactivity. Most notably, patients reached a normal level of suppression frequency. In addition, sup- pression frequency and electrocortical reactivity emerged as unidirec- tional predictors of subsequent anxiety while the role of other types of emotion regulation was less clear. This suggests that emotion regula- tion, and especially suppression, may play a complex, causal role in CBT for SAD.
If these findings are replicated, they might have important theore- tical and clinical implications for models of CBT for SAD. Models and protocols may need to assign a greater role to suppression and a lesser role to reappraisal. Our study also highlights the importance of multi- method designs when examining emotion regulation during treatment while specifying the exact nature of emotion dysregulation and focusing on the possible difficulties in implementing otherwise intact regulation abilities in daily life.
Author Contributions
Y.K. developed the concept, performed the data analysis and in- terpretation and drafted the paper for the current report under the supervision of J.D.H. J.D.H developed the overall study concept and design. All authors contributed to data collection, provided critical re- visions and approved the final manuscript for submission.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influ- ence the work reported in this paper.
Acknowledgements
The authors would like to thank Roni Pener-Tessler, Ariela Friedman, Asher Strauss, Yael Milgram, Shai Avishay, Michal Kovacs, and Maayan Langmass for their help in data collection and Jason Moser and members of the clinical psychophysiology lab for their help in analyses of the ERP data. Preparation of this manuscript was supported by a grant from the Israel Science Foundation (grant #332/09) to the
last author, the Sam and Helen Beber Chair of Clinical Psychology at The Hebrew University. The funding sources did not have any in- volvement in conducting the study and preparing the manuscript.
Supplementary materials
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.jad.2020.10.021.
References
American Psychiatric Association (2000). Diagnostic and statistical manual of mental dis- orders (4th ed., text rev. ed.). Washington, DC: Author.
Carthy, T., Horesh, N., Apter, A., Edge, M.D., Gross, J.J., 2010. Emotional reactivity and cognitive regulation in anxious children. Behaviour Research and Therapy 5 (48), 384–393.
Clark, D.M., 2005. A cognitive perspective on social phobia. In: Crozier, W.R., Alden, L.E. (Eds.), The Essential Handbook of Social Anxiety for Clinicians. Wiley, Chichester, pp. 193–218.
Cole, D., Maxwell, S., 2003. Testing mediational models with longitudinal data: questions and tips in the use of structural equation modeling. Journal of Abnormal Psychology 112 (4), 558–577.
Connor, K.M., Davidson, J.R., Churchill, L.E., Sherwood, A., Weisler, R.H., Foa, E., 2000. Psychometric properties of the Social Phobia Inventory (SPIN): New self-rating scale. The British Journal of Psychiatry 176 (4), 379–386.
Dryman, M.T., Heimberg, R.G., 2018. Emotion regulation in social anxiety and depres- sion: a systematic review of expressive suppression and cognitive reappraisal. Clinical Psychology Review 65, 17–42.
Falkenström, F., Finkel, S., Sandell, R., Rubel, J.A., Holmqvist, R., 2017. Dynamic models of individual change in psychotherapy process research. Journal of Consulting and Clinical Psychology 85 (6), 537–549.
Farmer, A.S., Kashdan, T.B., 2012. Social anxiety and emotion regulation in daily life: Spillover effects on positive and negative social events. Cognitive Behaviour Therapy 41 (2), 152–162.
Foti, D., Hajcak, G., 2008. Deconstructing reappraisal: Descriptions preceding arousing pictures modulate the subsequent neural response. Journal of Cognitive Neuroscience 20 (6), 977–988.
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 39 (4), 249–258.
Goldin, P.R., Lee, I., Ziv, M., Jazaieri, H., Heimberg, R.G., Gross, J.J., 2014a. Trajectories of change in emotion regulation and social anxiety during Cognitive Behavioral therapy for social anxiety disorder. Behaviour Research and Therapy 56, 7–15.
Goldin, P.R., Manber, T., Hakimi, S., Canli, T., Gross, J.J., 2009a. Neural bases of social anxiety disorder: emotional reactivity and cognitive regulation during social and physical threat. Archives of General Psychiatry 66 (2), 170–180.
Goldin, P.R., Manber-Ball, T., Werner, K., Heimberg, R., Gross, J.J., 2009b. Neural me- chanisms of cognitive reappraisal of negative self-beliefs in social anxiety disorder. Biological Psychiatry 66 (12), 1091–1099.
Goldin, P.R., Morrison, A.S., Jazaieri, H., Heimberg, R.G., Gross, J.J., 2017. Trajectories of social anxiety, cognitive reappraisal, and mindfulness during an RCT of CBGT versus MBSR for social anxiety disorder. Behaviour Research and Therapy 97, 1–13.
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., 2014b. 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.
Gross, J., John, O., 2003. Individual differences in two emotion regulation processes: Implications for affect, relationships, and well-being. Journal of personality and so- cial psychology 85 (2), 348–362.
Gross, J.J., 2015. Emotion regulation: Current status and future prospects. Psychological Inquiry 26 (1), 1–26.
Hajcak, G., Nieuwenhuis, S., 2006. Reappraisal modulates the electrocortical response to unpleasant pictures. Cognitive, Affective, & Behavioral Neuroscience 6 (4), 291–297.
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.
Hofmann, S.G., Curtiss, J.E., Hayes, S.C., 2020. Beyond linear mediation: Toward a dy- namic network approach to study treatment processes. Clinical Psychology Review 76.
Hofmann, S.G., Sawyer, A.T., Fang, A., Asnaani, A., 2012. Emotion dysregulation model of mood and anxiety disorders. Depression and Anxiety 29 (5), 409–416.
Huppert, J.D., Kivity, Y., Cohen, L., Strauss, A., Elizur, Y., Weiss, M., 2018. A pilot ran- domized clinical trial of cognitive behavioral therapy versus attentional bias mod- ification for social anxiety disorder: An examination of outcomes and theory-based mechanisms. Journal of Anxiety Disorders 59, 1–9.
Jaeger, B.C., Edwards, L.J., Das, K., Sen, P.K., 2016. An R 2 statistic for fixed effects in the
Y. Kivity, et al. Journal of Affective Disorders 279 (2021) 334–342
341
generalized linear mixed model. Journal of Applied Statistics 44 (6), 1086–1105. Jaeger, B. C., & R Core Team. (2016). r2glmm: R Squared for Mixed (Multilevel) Models,
R package version 0.1.1 [Computer software]. Retrieved from https://github.com/ bcjaeger/r2glmm.
Kashdan, T.B., Roberts, J.E., 2007. Social anxiety, depressive symptoms, and post-event rumination: Affective consequences and social contextual influences. Journal of Anxiety Disorders 21 (3), 284–301.
Kendall, P., Marrs-Garcia, A., Nath, S., Sheldrick, R., 1999. Normative comparisons for the evaluation of clinical significance. Journal of Consulting and Clinical Psychology 67 (3), 285–299.
Kinney, K.L., Burkhouse, K.L., Klumpp, H., 2019. Self-report and neurophysiological in- dicators of emotion processing and regulation in social anxiety disorder. Biological Psychology, 142 126–131.
Kivity, Y., Huppert, J.D., 2018. Are individuals with social anxiety successful in reg- ulating their emotions? A mixed-method investigation using self-report, behavioral, and event-related potentials measures. Journal of Affective Disorders 236, 298–305.
Kivity, Y., Huppert, J.D., 2019. Emotion regulation in social anxiety: A systematic in- vestigation using self-report, behavioral, and event-related potentials measures. Cognition & Emotion 33, 213–230.
Kivity, Y., Huppert, J.D., 2016. Does cognitive reappraisal reduce anxiety? A daily diary study of a micro-intervention with individuals with high social anxiety. Journal of Consulting and Clinical Psychology 84 (3), 269–283.
Kocovski, N.L., Fleming, J.E., Hawley, L.L., Ho, M.-H.R., Antony, M.M., 2015. Mindfulness and acceptance-based group therapy and traditional cognitive beha- vioral group therapy for social anxiety disorder: Mechanisms of change. Behaviour Research and Therapy 70, 11–22.
Kocovski, N.L., Fleming, J.E., Hawley, L.L., Huta, V., Antony, M.M., 2013. Mindfulness and acceptance-based group therapy versus traditional cognitive behavioral group therapy for social anxiety disorder: A randomized controlled trial. Behaviour Research and Therapy 12 (51), 889–898.
Lang, P.J., Bradley, M.M., Cuthbert, B.N., 2008. International affective picture system (IAPS): Affective ratings of pictures and instruction manual. University of Florida, Gainesville, FL.
Levin, J.B., Marom, S., Gur, S., Wechter, D., Hermesh, H., 2002. Psychometric properties and three proposed subscales of a self-report version of the Liebowitz Social Anxiety Scale translated into Hebrew. Depression and Anxiety 16 (4), 143–151.
Liebowitz, M.R., 1987. Social phobia. Modern problems of Pharmacopsychiatry 22, 141–173.
Moran, T.P., Jendrusina, A.A., Moser, J.S., 2013. The psychometric properties of the late
positive potential during emotion processing and regulation. Brain Research 1516, 66–75.
Morrison, A.S., Heimberg, R.G., 2013. Social anxiety and social anxiety disorder. Annual Review of Clinical Psychology 9, 249–274.
Moscovitch, D.A., 2009. What is the core fear in social phobia? A new model to facilitate individualized case conceptualization and treatment. Cognitive and Behavioral Practice 16, 123–134.
Moscovitch, D.A., Gavric, D.L., Senn, J.M., Santesso, D.L., Miskovic, V., Schmidt, L.A., Antony, M.M., 2012. Changes in judgment biases and use of emotion regulation strategies during Cognitive Behavioral therapy for social anxiety disorder: Distinguishing treatment responders from nonresponders. Cognitive Therapy and Research 36 (4), 261–271.
Moser, J.S., Hartwig, R., Moran, T.P., Jendrusina, A.A., Kross, E., 2014. Neural markers of positive reappraisal and their associations with trait reappraisal and worry. Journal of Abnormal Psychology 123 (1), 91–105.
Nock, M.K., 2007. Conceptual and design essentials for evaluating mechanisms of change. Alcoholism: Clinical and Experimental Research 31 (S3), 4S–12S.
Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., & R Core Team. (2016). nlme: Linear and Nonlinear Mixed Effects Models, R package version 3.1-128 [Computer software]. Retrieved from http://CRAN.R-project.org/package=nlme.
Proudfit, G.H., Dunning, J., Foti, D., Weinberg, A., 2013. Temporal dynamics of emotion regulation. In: Gross, J.J. (Ed.), Handbook of Emotion Regulation, 2nd ed. Guilford Publications, New York, pp. 43–57.
Roth-Ledley, D.A., Foa, E.B., Huppert, J.D., 2006. Comprehensive Cognitive Behavior Therapy for Social Phobia: A Treatment Manual. University of Pennsylvania, Philadelphia, PA.
Sheehan, D.V., Lecrubier, Y., Sheehan, K.H., Amorim, P., Janavs, J., Weiller, E., Dunbar, G.C., 1998. The Mini-International Neuropsychiatric Interview (MINI): The devel- opment and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. Journal of Clinical Psychiatry 59, 22–33.
Strauss, A.Y., Kivity, Y., Huppert, J.D., 2019. Emotion Regulation Strategies in Cognitive Behavioral Therapy for Panic Disorder. Behavior Therapy 50, 659–671.
Wang, L., Maxwell, S., 2015. On disaggregating between-person and within-person effects with longitudinal data using multilevel models. Psychological Methods 20 (1), 63–83.
Yuan, L., Zhou, R., Hu, S., 2014. Cognitive reappraisal of facial expressions: Electrophysiological evidence of social anxiety. Neuroscience Letters 577, 45–50.
Ziv, M., Goldin, P.R., Jazaieri, H., Hahn, K.S., Gross, J.J., 2013. Emotion regulation in social anxiety disorder: behavioral and neural responses to three socio-emotional tasks. Biology of Mood & Anxiety Disorders 3 (1), 20–36.
Y. Kivity, et al. Journal of Affective Disorders 279 (2021) 334–342
342
- The role of expressive suppression and cognitive reappraisal in cognitive behavioral therapy for social anxiety disorder: A study of self-report, subjective, and electrocortical measures
- 1 Introduction
- 2 Method
- 2.1 Participants
- 2.2 Measures
- 2.3 Treatment and therapists
- 2.4 Procedure
- 2.5 Data Analyses
- 3 Results
- 3.1 Changes in regulation (H1) and equivalency to HCs (H3)
- 3.1.1 Change in self-reports
- 3.1.2 Change in lab-based measures
- 3.1.3 Equivalency to HCs
- 3.2 Cross-lagged associations between emotion regulation and anxiety (Hypothesis 2)
- 3.2.1 Self-reported emotion regulation
- 3.2 2. Lab-based emotion regulation
- 4 Discussion
- 4.1 Changes in emotion-regulation during treatment
- 4.2 The association between emotion-regulation and anxiety
- 5 Limitations
- 6 Conclusions
- Author Contributions
- Declaration of Competing Interest
- Acknowledgements
- Supplementary materials
- References