Boston University OpenBU

http://open.bu.edu

Theses & Dissertations

Boston University Theses & Dissertations

2015

Cognitive control Training as an adjunct to behavioral activation therapy in the treatment of depression Moshier, Samantha J. https://hdl.handle.net/2144/14051 Boston University

BOSTON UNIVERSITY GRADUATE SCHOOL OF ARTS AND SCIENCES

Dissertation

COGNITIVE CONTROL TRAINING AS AN ADJUNCT TO BEHAVIORAL ACTIVATION THERAPY IN THE TREATMENT OF DEPRESSION

by

SAMANTHA MOSHIER B.A., University of Pennsylvania, 2007 M.A., Boston University, 2010

Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy 2015

© 2015 SAMANTHA MOSHIER All rights reserved

Approved by

First Reader

_________________________________________________________ Michael Otto, Ph.D. Professor of Psychological & Brain Sciences

Second Reader

_________________________________________________________ Heather Murray Latin, Ph.D. Research Assistant Professor of Psychological & Brain Sciences

Third Reader

_________________________________________________________ Stefan Hofmann, Ph.D. Professor of Psychological & Brain Sciences

  ACKNOWLEDGMENTS Thank you to my dissertation committee: Dr. Michael Otto, Dr. Heather Murray, Dr. Stefan Hofmann, Dr. Todd Farchione, and Dr. Lisa Smith. I have been fortunate to have you all not only as committee members but also as supervisors and teachers over the past six years, and I thank you for your outstanding professional guidance. To Dr. Michael Otto, thank you for your steady encouragement and generous mentorship. You have taught me the intellectual stuff – thinking critically and writing well and all of that - but perhaps as importantly, you’ve been a model for optimistically pursuing your research ideas and having fun doing science. Many thanks to the Translational Research Program lab members who served as evaluators and therapists in this project: Xandra Kredlow, Kristin Szuhany, Tetsuhiro Yamada, Mateo Bugatti, Anne Ward, Elizabeth Kaiser, and Margot Iverson. And special thanks to Bridget Hearon, Amanda Calkins, Kristin Szuhany, Xandra Kredlow, and Angela Fang, who made graduate school more fun that I thought it ever would be. This dissertation is dedicated with love to my family: Tim McInerny and Karen, Kevin, Michaela, and Juliana Moshier. Tim, thank you for your constant love and support and for putting up with years of psychology acronyms. I promise I’ll have a real job soon.

 

iv

  COGNITIVE CONTROL TRAINING AS AN ADJUNCT TO BEHAVIORAL ACTIVATION THERAPY IN THE TREATMENT OF DEPRESSION (Order No.

)

SAMANTHA MOSHIER Boston University Graduate School of Arts and Sciences, 2015 Major Professor: Michael W. Otto, Professor of Psychological & Brain Sciences ABSTRACT Major depressive disorder (MDD) is characterized by reduced activation of the dorsolateral prefrontal cortex (DLPFC), a brain region involved in both emotion regulation and basic cognitive control processes. Recent studies have indicated that computerized interventions designed to activate the DLPFC can reduce depressive symptoms. The current study was a randomized controlled trial which extends this research to test whether one such program, called Cognitive Control Training (CCT), enhances depression treatment outcomes when used in adjunct to brief behavioral activation therapy for depression (BATD), an empirically-supported outpatient intervention. This study also explored whether the effects of BATD + CCT treatment on depression were mediated by changes in rumination and cognitive control. In a sample of thirty-four adults diagnosed with MDD, participants were randomly assigned to complete four sessions of either computerized CCT or a non-active computerized control task, concurrently with four sessions of BATD. Completion of the assigned computerized task took place immediately before each of the four BATD therapy sessions. Depression symptoms and proposed treatment mediators were assessed at baseline, mid-treatment,  

v

  post-treatment, and four-week follow-up visits. I hypothesized that compared to the control group, participants receiving adjunctive CCT would demonstrate significantly reduced depressive symptoms. I also hypothesized that these effects would be mediated by changes in inhibitory control and set-shifting performance in the context of negative emotional material, as well as by changes in ruminative brooding. Results did not support these hypotheses. Depressive symptoms were reduced over time in both treatment conditions, with no significant difference between treatment conditions. Assignment to CCT was not associated with changes in the proposed mediators. Furthermore, exploratory analyses found minimal evidence that performance on inhibitory control and set-shifting tasks were related to baseline clinical characteristics (such as depression severity, rumination, or anxiety symptoms) or treatment outcomes. The results of this study support the potential for BATD as a brief, low-cost, flexible intervention for the treatment of depression and further show that CCT administered in adjunct to a 4-session BATD program does not add clinical benefit in the treatment of depression. This study and other recent research suggest that the effects of CCT may not be as robust as previously indicated, highlighting the need for continued investigation of the conditions under which CCT may be effective.

 

vi

  TABLE OF CONTENTS

ACKNOWLEDGMENTS ................................................................................................. iv   TABLE OF CONTENTS.................................................................................................. vii   LIST OF TABLES ........................................................................................................... viii   LIST OF FIGURES ........................................................................................................... ix   LIST OF ABBREVIATIONS ............................................................................................. x   INTRODUCTION .............................................................................................................. 1   METHODS ....................................................................................................................... 14   RESULTS ......................................................................................................................... 32   DISCUSSION ................................................................................................................... 48   CONCLUSION ................................................................................................................. 66   TABLES ........................................................................................................................... 67   FIGURES .......................................................................................................................... 80   REFERENCES ................................................................................................................. 84   CURRICULUM VITAE ................................................................................................... 97    

 

vii

  LIST OF TABLES Table 1. Study measures and procedures by visit. ........................................................... 67   Table 2. Participant characteristics of the Intent-to-Treat sample. ................................ 68   Table 3. Participant characteristics of the completer sample. ........................................ 69   Table 4. Participant characteristics by completer and dropout status. ........................... 70   Table 5. Primary and secondary outcome measures for all randomized participants at baseline, mid-treatment, post-treatment, and follow-up (N = 34). ........................... 71   Table 6. Primary and secondary outcome measures for all treatment completers at baseline, mid-treatment, post-treatment, and follow-up (n = 26). ........................... 72   Table 7. Response and remission rates by treatment condition....................................... 73   Table 8. Means and Standard Deviations (SDs) for NAP task response times and NAP scores over time. ....................................................................................................... 74   Table 9. Hierarchical multiple regression analyses of the predictive influence of baseline NAP-negative scores and treatment condition on treatment outcomes at follow-up.75   Table 10. Means and standard deviations for response time on Internal Shift Task (IST) Trials ......................................................................................................................... 77   Table 11. Hierarchical multiple regression analyses of the predictive influence of baseline IST emotion switch cost and treatment condition on week 8 treatment outcomes. .................................................................................................................. 78    

 

viii

  LIST OF FIGURES Figure 1. CONSORT diagram of participant enrollment. ............................................... 80   Figure 2. Negative Affective Priming Task: Example of a priming trial (in the negative information condition) .............................................................................................. 81   Figure 3. Negative Affective Priming Task: Example of a control trial. ......................... 82   Figure 4. NAP-Positive Scores over time by treatment condition. .................................. 83    

 

ix

  LIST OF ABBREVIATIONS ADIS

Anxiety Disorder Interview Schedule for DSM-IV-TR

ANOVA

Analysis of Variance

APA

American Psychological Association

BA

Behavioral Activation

BAI

Beck Anxiety Inventory

BATD

Brief Behavioral Activation Therapy for Depression

BDI

Beck Depression Inventory – II

CCT

Cognitive Control Training

DLPFC

Dorsolateral Prefrontal Cortex

DSM-IV

Diagnostic and Statistical Manual of Mental Disorder, fourth edition

DSM-5

Diagnostic and Statistical Manual of Mental Disorder, fifth edition

fMRI

Functional Magnetic Resonance Imaging

ISI

Inter-stimulus Interval

IST

Internal Shift Task

ITT

Intent to Treat

LGM

Latent Growth Modeling

MADRS

Montgomery-Asberg Depression Rating Scale

MDD

Major Depressive Disorder

NAP

Negative Affective Priming

PASAT

Paced Auditory Serial Addition Task

PVT

Peripheral Vision Task

 

x

  RRS

Ruminative Response Scale

RPI

Reward Probability Index

RPI-ES

Reward Probability Index – Environmental Suppressor subscale

RPI-RP

Reward Probability Index – Reward Probability subscale

RT

Response Time

SCID

Structured Clinical Interview for DSM-IV

TAU

Treatment as usual

tDCS

Trans-cranial Direct Current Stimulation

 

xi

 

1 INTRODUCTION Major Depressive Disorder (MDD) is a highly prevalent disorder that places

significant burden on society and those suffering from the illness. Individuals who struggle with depression face significant disability including reduced rates of employment and productivity (Stewart, Ricci, Chee, Hahn, & Morganstein, 2003), increased risk of mortality (Rovner et al., 1991) and suicide (Chen & Dilsaver, 1996), social stigma (Cooper, Corrigan, & Watson, 2003; Corrigan, Markowitz, Watson, Rowan, & Kubiak, 2003), and interpersonal problems (Burns, Sayers, & Moras, 1994; Gotlib & Whiffen, 1989). The course of depression is relapsing, and each episode further denotes an increased risk for another depressive episode (Solomon et al., 2000). Furthermore, residual symptoms following treatment are predictive of relapse, making full, sustained recovery an important target for treatments (Cornwall & Scott, 1997; Fava, Ruini, & Belaise, 2006). Pharmacotherapy and psychosocial interventions such as cognitive therapy and behavioral activation therapy have shown efficacy in the treatment of depression, and the recent success of brief behavioral activation protocols across depressed populations (e.g., Lejuez, Hopko, Acierno, Daughters, & Pagoto, 2011) have shown particular promise for increasing efficiency and access to treatment. However, one concern is that the value of such interventions may be derailed by the ongoing negative ruminative style common to patients struggling with depression. MDD is characterized by excessive rumination and difficulties disengaging from negative emotional material (for reviews, see Gotlib & Joormann, 2010 and NolenHoeksema, Wisco, & Lyubomirsky, 2008). For instance, a patient with depression may  

 

2

overly attend to the meaning and the consequences of his or her emotional experience (“My mood is so low” or “How did things get this bad?”). This repetitive yet passive focus on one’s inner experience leads individuals to recall more negative memories, interpret current situations in a negative way, and to have more pessimistic beliefs about the future (Nolen-Hoeksema, 2000). Research suggests that self-focused, ruminative response styles are associated with poorer treatment outcome in MDD (Ciesla & Roberts, 2002; Schmaling et al., 2002). Furthermore, ruminative response styles are predictive of the onset of a major depressive episode (Abela & Hankin, 2011), longer episodes of depression (Abela & Hankin, 2011; Nolen-Hoeksema, Morrow, & Frederickson, 1993), suicide ideation (Miranda & Nolen-Hoeksema, 2007), and relapse following treatment (Michalak, Holz, & Teismann, 2011). Closer examination of the construct of rumination has led to differentiation between ruminative brooding, defined as a more passive selffocus, and ruminative reflection, defined as a more purposeful and constructive process (Treynor, Gonzalez, & Nolen-Hoeksema, 2003). Depression has been more strongly associated with brooding rather than reflection (Joormann, Dkane, & Gotlib, 2006). The clinical features of depression have also been associated with deficits in cognitive control abilities such as attentional and inhibitory control. For example, deficits in inhibiting negative emotional information from entering working memory are indicated by a number of studies involving negative affective priming tasks. In these tasks, participants are asked to respond to a target stimulus in the presence of an irrelevant emotional distractor (i.e., a negative emotional word or face). On a subsequent trial, the distractor stimulus becomes the target, and the participant’s delay in response  

 

3

latency is considered to reflect the strength of the negative priming effect. A number of studies using this type of task have indicated that compared with healthy participants, dysphoric and depressed individuals exhibit reduced inhibitory capacity for negative emotional information (Goeleven, DeRaedt, Baert, & Koster, 2006; Joormann, 2004; Joormann, 2006; Joormann & Gotlib, 2010). Joormann and Gotlib (2010) also found that reduced inhibitory control of negative material was associated with a shift toward less adaptive emotion regulation strategies such as increased use of suppression and reduced use of reappraisal and reflection. Furthermore, depressed individuals have been found to have greater difficulty both removing and manipulating negative information from working memory compared with control participants (Joormann & Gotlib, 2008; Joormann, Levens, & Gotlib, 2011). Accordingly, Holtzheimer and Mayberg (2011) have suggested that the defining feature of depression is not the mere presence of negative mood, but rather, the tendency to get “stuck in a rut”; that is, to have a tendency to inappropriately enter into and inability to disengage with a depressed mood state. Consistent with this, a recent study of depressed individuals indicated an impaired ability to switch between mental representations in working memory in response to facial expressions (De Lissnyder et al., 2012). Importantly, the extent of these deficits in inhibiting, manipulating, and removing negative emotional information from working memory has been positively associated with increased rumination (Joormann, 2006; Joormann & Tran, 2009; De Lissnyder et al., 2012). Additionally, rumination has been associated with increased attentional bias for negative words, even when controlling for severity of depressive symptoms  

 

4

(Donaldson, Lam, & Mathews, 2007). Rumination also appears to disrupt problem solving ability. For example, individuals engaged in a ruminative task, who were then instructed to problem-solve, generated lower quality solutions compared to those who had previously engaged in a distracting task (Lyubomirsky & Nolen-Hoeksema, 1995). Set-shifting ability, another important component of executive control, is also compromised. For example, individuals with higher levels of rumination (Davis & Nolen-Hoeksema, 2000), and clinically depressed individuals (Harvey et al., 2004; Merriam, Thase, Haas, Keshavan, & Sweeney, 1999) commit more perseverative errors on the Wisconsin Card Sorting Task than healthy controls. There is evidence that the impairments in cognitive control and emotion regulation common to depression involve abnormalities in activation and connectivity within limbic-cortical brain regions. Neurobiological models, such as those proposed by Mayberg (Mayberg, 1997; Mayberg et al., 1999) and Phillips (Phillips, Drevets, Rauch, & Lane, 2003), suggest that depression is characterized by dysregulation of two systems: (1) a ventral pathway, which is responsible for the identification of emotional stimuli and the production of affective states and behaviors and includes the amygdala and limbic regions, and (2) a dorsal pathway, which is involved in the effortful control of emotions, and includes the prefrontal cortex, hippocampus, and dorsal anterior cingulate gyrus. Concerning the first pathway, patients with depression demonstrate hyperactivity in the amygdala, thalamus, and ventral regions of the limbic system (Drevets et al., 1992; Mayberg et al., 1999) and increased activity in the amygdala following exposure to negative emotional stimuli (Siegle, Steinhauer, Thase, Stenger, & Carter, 2002; Siegle,  

 

5

Thompson, Carter, Steinhauer, & Thase, 2007). Concerning the second pathway, depressed patients demonstrate reduced activation of the dorsolateral prefrontal cortex (DLPFC; e.g., Davidson, 1994; Drevets, 1999; Mayberg et al., 1999; Siegle et al., 2007), the region of the brain responsible for the effortful regulation of emotion (e.g., Ochsner et al., 2004; Ochsner & Gross, 2008) as well as basic executive control processes such as working memory (for review, see Smith & Jonides, 1999). Recovery from depression has been associated with increased DLPFC activation (Fales et al., 2009; Mayberg et al., 2000) and a reduction in amygdala hyperactivation (Mayberg et al., 2000; Sheline et al., 2001), suggesting that these disruptions are state-specific. Taken together, these findings suggest that depression is maintained by “bottom-up” maladaptive activation of ventral (and more specifically, limbic) regions, which leads to increased emotional reactivity and the presence of negative mood, and “top-down” failure of the dorsal pathway, that is, failure to recruit prefrontal control to attenuate this response (Philips et al. 2003). Cognitive Control Training Given these neuropsychological findings, Siegle (1999) has proposed that enhancement of cognitive control might alleviate depressive symptoms by enhancing topdown regulation of negative affect. Consistent with this hypothesis, existing empirically based psychological treatments for depression use techniques that may increase cognitive control. For instance, cognitive therapy for depression encourages patients to override their automatic negative thought processes and instead identify alternative thoughts. Mindfulness-based therapies utilize attentional training, such as exercises in which patients practice redirecting attention away from ruminative thoughts and toward the  

 

6

present moment. However, it remains unclear whether these interventions exert their effects through increased cognitive control. In order to test more directly whether increasing cognitive control might reduce depression, Siegle and colleagues (2007) developed a computerized neurobehavioral therapy that directly targets cognitive control processes such as working memory and attention. Called Cognitive Control Training (CCT), this program consists of two computerized training tasks designed to activate and strengthen the prefrontal cortex with the goal of correcting the hypoactivity apparent in these areas during depressed mood states. The first task, the Wells training task, was designed to improve selective attention to specific information, and perhaps thereby increase ability to selectively attend away from ruminative thoughts (Wells, 2000). The Wells task asks participants to listen to an array of naturalistic sounds and to attend to one sound at a time, switch attention between sounds, and count sounds (Siegle et al., 2014). The second task is a variant of the Paced Auditory Serial Addition Task (PASAT; Gronwall, 1977), which is known to specifically activate the prefrontal cortex (Lazeron, Rombouts, de Sonneville, Barkhof, & Scheltens, 2003). In this version of the PASAT, participants must continuously add serially presented digits, thereby requiring holding information in working memory. In combination, one session of CCT takes approximately 30 minutes to complete. Several studies have now shown that CCT significantly reduces depressive symptoms. In the first study by Siegle and colleagues (2007), 31 severely depressed patients participating in an intensive outpatient program were randomized to receive treatment as usual (medication management, group therapy based on dialectical behavior  

 

7

therapy principles, and milieu therapy) or treatment as usual plus six sessions of CCT over a two-week period. At post-treatment, participants in the CCT condition demonstrated significantly reduced scores on the Beck Depression Inventory (d = 1.28) and the Response Styles Questionnaire, a self-report measure of rumination (d = 1.26). Moreover, six participants in the CCT condition completed fMRI assessment at pre- and post- treatment. Results showed that following CCT, these participants demonstrated increased DLPFC activity on a novel cognitive task, as well as reduced disruption in amygdala activity during an emotional responding task. These findings suggest that CCT can effectively reduce symptom severity in severely depressed patients and provide initial support for the validity of CCT’s proposed mechanisms of change. In a second study, Calkins and colleagues (2014) examined the effects of CCT in a community sample of depressed adults. Forty-eight non-treatment seeking adults with depressed mood (as defined by a BDI score ≥ 17) were randomized to 3 sessions of CCT or a control task (peripheral vision training, or PVT) over a 2-week period. Participants who received CCT experienced a mean reduction of 6 points on the BDI, a significant difference (d = 0.73) from the comparison group, whose mean reduction was less than 1 point. These results suggest that the benefits of CCT generalize to less severely depressed samples and may emerge after a low dose. Recently, two studies have examined CCT in combination with transcranial direct current stimulation (tDCS), a form of neurostimulation that, when applied to the left DLPFC, shows positive yet modest effects on depression symptom severity. Both studies provide more modest estimates of the benefit of CCT in reducing depressive symptoms.  

 

8

In the first study, Segrave and colleagues (2013) compared the efficacy of three treatment conditions: tDCS + CCT, sham tDCS + CCT, and tDCS + sham CCT. Treatment sessions occurred daily for 5 days; following this period all groups demonstrated significant reductions in depression symptoms. At a three-week follow-up, only the tDCS + CCT group demonstrated a sustained effect, suggesting that CCT alone did not lead to sustained gains. This is important because despite the positive effects demonstrated in Calkins et al. (2014), the study was limited by the lack of a follow-up assessment period. A second tDCS study compared depressed participants receiving 10 sessions of CCT + tDCS vs. CCT + sham tDCS over a 2-week period (Brunoni et al., 2014). Both groups of participants demonstrated clinical improvement at the end of treatment, however, the response rate was quite low in both conditions (25% across groups). Exploratory analyses indicated that CCT was more effective for older participants and participants who performed better on the CCT task throughout the study (Brunoni, et al., 2014). These data show some promise for CCT as a mood enhancing intervention but suggest that continued study of the intervention is necessary. It is particularly important to continue to investigate the effects of CCT across a range of clinical contexts. Although CCT was designed as an adjunctive intervention, it has only been tested in adjunct to intensive outpatient (pharmacologic and psychosocial; Siegle, Ghinassi, & Thase, 2007) or biological (tDCS; Segrave et al., 2013; Brunoni et al., 2014) treatment for depression. One question of particular interest, then, is whether CCT might enhance treatment outcomes when used in adjunct to an empirically-supported psychosocial  

 

9

intervention delivered to outpatients, boosting the efficacy of treatment in an efficient, low cost, and transportable manner. A second question pertains to the mechanisms of change of CCT. Siegle’s initial investigation of CCT showed that it led to apparent normalization of activity in the left amygdala and left DLPFC during an emotional task. However, this analysis was only conducted in six individuals who received CCT, and the lack of control group and small sample size limits the conclusions that can be drawn. In addition, it remains unclear if CCT enhances cognitive control, and if so, what specific cognitive processes are affected. Siegle and colleagues (2007) attempted to assess this question using a digit span task, but found that all participants performed at ceiling prior to receiving CCT. Given that depressed participants have tended to show impaired inhibitory control and set-shifting performance specifically in the context of emotional stimuli, these variables may be mediators of CCT effects on depression symptoms. It is also possible that by enhancing cognitive control, CCT allows for improved engagement in psychosocial treatment. For example, Addis and Carpenter (1999) found that higher levels of rumination were associated with less positive reactions to action-oriented depression treatment rationales. Accordingly, in the present study I examine whether adjunctive treatment with CCT results in differential improvement in inhibitory control and set-shifting, reductions in rumination, and whether these changes are associated with enhanced treatment outcome as well as homework adherence. Behavioral Activation Treatment for Depression

 

 

10 The current study examines the effect of CCT in adjunct to brief behavioral

activation treatment for depression (BATD). Behavioral activation (BA) therapy developed out of the behavioral models of depression postulated by Ferster (1973) and Lewinsohn (1974) which suggested that depression is maintained by a low rate of behavior, which itself is maintained by a lack of positive reinforcement from the environment. From this theoretical basis, researchers began to develop interventions involving the monitoring and scheduling of reinforcing activities, referring to these strategies as activity scheduling. Several early studies found that activity scheduling reduced depression (Lewinsohn, Sullivan, & Grosscup, 1980; Barrera 1979), resulting in the inclusion of activity scheduling in a number of treatment packages for depression, including Beck’s cognitive therapy and the Control your Depression protocol (Beck, 1979; Lewinsohn, Munoz, Youngren, & Zeiss, 1978). Yet attention to these strategies as a stand-alone treatment was modest throughout the 1980s and early 1990s. This changed following a seminal study by Jacobson and colleagues (1996) that compared a full course of cognitive therapy to activity scheduling (given the label of behavioral activation) and automatic thought retraining alone. Results revealed that behavioral activation alone was equally efficacious to the complete cognitive therapy treatment package. Since this time, interest in behavioral activation as a stand-alone treatment has grown significantly (for review, see Dimidjian, Barrerra, Martell, Munoz, & Lewinsohn, 2011). Behavioral activation protocols for depression have been developed and tested by a number of different research groups; although these protocols differ in a number of ways, they are alike in that they are grounded in behavioral theories of depression and  

 

11

exclusively target behavior change. Treatment is structured and goal-directed, and aims to engage patients in adaptive, rewarding activities and to reduce involvement in behaviors that maintain depressive symptoms (Dimidjian et al., 2011). To this end, primary treatment strategies include self-monitoring of daily activities and mood, scheduling of activities that bring patients a sense of pleasure or mastery, and identifying and reducing avoidance behaviors that increase depressive symptoms. Several BA protocols have been shown to be efficacious in the treatment of severe depression, with one large randomized controlled trial finding that a 24-session protocol performed comparably to antidepressant medication and outperformed cognitive therapy of a similar length (Dimidjian et al., 2006). BA is considered to have “strong research support” by the APA Presidential Task Force on Evidence-Based Practice (2006), and has demonstrated clinically significant positive effects across a range of clinical settings and treatment groups, including adult and elderly patients, psychiatric inpatients, depressed women with breast cancer, cigarette smokers, and illicit drug users (Daughters et al., 2008, Dimidjian et al., 2006; Hopko, Lejuez, LePage, Hopko, & McNeil, 2003; Hopko, Bell, Armento, Hunt, & Lejuez, 2005; Macpherson et al., 2010). The current study makes use of a specific BA protocol developed by Lejuez and colleagues (2011) called the Brief Behavioral Activation Treatment for Depression (BATD). This time-limited protocol is limited to treatment components related to behavioral activation and focuses on monitoring and scheduling activities within a values-based framework (Lejuez, Hopko, Acierno, Daughters, & Pagoto, 2011). Several trials provide support for the efficacy of BATD. Hopko and colleagues (2003) found that  

 

12

compared to treatment as usual, BATD significantly reduced depression symptom severity in an inpatient psychiatric sample. A more recent randomized controlled trial demonstrated that the BATD protocol significantly reduced depression and increased quality of life and social support among breast cancer patients with major depressive disorder, and that these clinical gains were maintained over a twelve-month period (Hopko et al., 2011). A randomized clinical trial examining the addition of BATD to standard tobacco cessation techniques in mildly depressed smokers found that the addition of BATD resulted in significantly reduced depressive symptoms and enhanced smoking abstinence (MacPherson et al., 2010). One of the strengths of the BATD protocol is that it is adaptable to various treatment lengths and has shown clinical benefit in very brief protocols. For instance, Daughters and colleagues (2008) found that a six-session BATD treatment for illicit drug users significantly reduced depressive symptoms. More recently, a compacted single session of BATD followed by two weeks of activity assignments was found to have a large effect (d = 1.61) on the reduction of depressive symptoms in undergraduate students compared with a no-treatment control condition (Gawrysiak, Nicholas, & Hopko, 2009). Despite the empirical support for BATD and its clinical advantages such as brevity and flexibility, there is room for improvement using this approach. Even in the study with the highest response rates to a full-length BA treatment, 25% of patients did not respond to treatment (Dimidjian et al., 2006). There is likely even greater need to improve on brief BATD treatments, which have been found to significantly reduce

 

 

13

depressive symptoms but typically result in smaller changes compared to lengthier treatments (Daughters et al., 2008; Magidson et al., 2011; Dimidjian et al., 2006). In sum, the reliable evidence for the efficacy of BATD, as well as its brevity and flexibility, make it an ideal platform from which to study the additive effects of CCT. BATD is a parsimonious treatment that can be effectively applied by non-specialists with little training in psychotherapy (Ekers, Richards, McMillan, Bland, & Gilbody, 2011). In addition, there is research suggesting that it may be amenable to computerized or other non-traditional treatment formats (Dimidjian et al., 2011), and recent trials of Internetbased BA packages show initial support for their feasibility and efficacy (Carlbring et al, 2013; O’Mahen et al., 2013). BATD is a primarily behavioral intervention, providing an appropriate backdrop in which the added effects of the cognitively focused CCT program may be evaluated. The addition of CCT as an adjunct to brief BATD provides an easily transportable cognitive component that is not included in the behaviorally-focused BATD protocol. The Current Study This investigation expands upon previous trials of CCT for depression by testing its efficacy in adjunct to BATD and by examining proposed mediators of treatment. Twenty-six individuals with a primary diagnosis of MDD received 4 sessions of BATD and were randomly assigned to concurrently complete either four sessions of CCT or four sessions of a computerized control condition called peripheral vision training (PVT). Depression symptoms and hypothesized mediators of treatment outcome were assessed at

 

 

14

pre-, mid-, and post-treatment and at a four-week follow-up. The aims and hypotheses were as follows: Aim 1, Hypothesis 1. To test whether adjunctive CCT adds significant clinical benefit to BATD. I hypothesized that relative to those in the BATD + PVT condition, individuals in the BATD + CCT condition would demonstrate greater reduction in depression symptoms and that this effect would be maintained over the four-week followup period. Aim 2, Hypothesis 2. To examine changes in ruminative brooding, cognitive control of emotional material, and homework compliance as potential mediators of response to CCT. I hypothesized that relative to the BATD + PVT condition, the BATD + CCT condition would be associated with reductions in rumination and increased cognitive control. Further, I hypothesized that these variables would mediate the relationship between treatment condition and treatment outcome and the relationship between treatment condition and BATD homework adherence. METHODS Design Participants received four weekly sessions of BATD and were randomly assigned to concurrently receive four sessions of CCT or four sessions of PVT during the same period. Treatment outcome was assessed by self-report and clinician-rated depressive symptom rating scales, with secondary outcomes of self-reported rumination and anxiety symptoms. Cognitive control (specifically, inhibitory control and set-shifting), environmental reward, and homework adherence were assessed at baseline, midpoint,  

 

15

post-treatment, and 4-week follow-up in order to evaluate potential mediators of treatment response. The design was double-blind, such that participants and the independent evaluators assessing depression severity were not informed of the randomized portion (CCT vs. PVT) of the treatment. The trial was registered with clinicaltrials.gov with identifier NCT01694719. Participant Enrollment Participants were recruited between October of 2012 and April of 2014 through community advertisements or were referred from the Center for Anxiety and Related Disorders at Boston University. Potential participants were screened via telephone interview before being invited to an in-person screening visit. The inclusion/exclusion criteria were as follows: Inclusion Criteria: 1. Between 18 and 65 years of age 2. Primary psychiatric diagnosis of major depressive disorder 3. Ability to read and speak English sufficiently to complete study procedures 4. If taking antidepressant or anxiolytic medication, participants must be taking a stabilized for a minimum period of at least 8 weeks prior to entry into the study 5. Willingness and ability to comply with the requirements of the study protocol Exclusion Criteria: 1. Lifetime history of bipolar disorder or psychotic disorder as assessed by SCID-IV or ADIS-IV-L

 

 

16 2. Neurological disorder such as Parkinson’s disease or traumatic brain injury as assessed by patient self report during the phone screen and again during the screening visit 3. Alcohol or substance dependence within the past 6 months as assessed by SCID or ADIS-IV-L 4. Substantial suicide risk, as indicated by a rating of 2 or greater on the suicide item of the BDI 5. Concurrent psychotherapy initiated within 2 months of baseline, or ongoing psychotherapy of any duration directed specifically toward treatment of the depression other than general supportive therapy 6. Current use of antipsychotics, stimulants, or modafinil Figure 1 provides a CONSORT diagram representing the flow of participants

through the trial. Following telephone pre-screening interviews, 43 participants attended the screening visit in which they provided informed consent and were administered the SCID-IV to assess eligibility. Six individuals did not meet criteria for study entry (3 did not meet diagnostic criteria for MDD, 2 had bipolar disorder, and 1 did not have a principal diagnosis of MDD). Of the 37 participants deemed eligible after the screening visit, 3 were lost to follow up prior to randomization and initiation of treatment. Randomization The remaining 34 participants were randomly assigned to CCT or PVT conditions. The randomization sequence was generated by the author using a computerized random number generator service. Randomization was stratified by  

 

17

severity of depression, where a BDI score greater than 29 was considered severe (consistent with the clinical severity ranges of the BDI; Beck, Steer, & Brown, 1996). Due to a disproportionate attrition rate (a greater number of individuals in the CCT condition failed to complete the study), the randomization schedule was modified halfway through the trial in order to ensure an equal number of completers in each condition. Measures Diagnosis. The Structured Clinical Interview for Axis I Disorders for DSM-IV (SCID-IV; First, Spitzer, Gibbon, & Williams, 2002) was used to assess the presence of MDD at the initial study visit and to rule out other diagnoses that would preclude participation in the study, such as a psychotic disorder or bipolar disorder. Individuals referred through the Center for Anxiety and Related Stress Disorders had already received a diagnostic interview using the Anxiety Disorders Interview Schedule for DSM-IV, Lifetime Edition (ADIS-IV-L; Brown, DiNardo & Barlow, 1994); therefore, the depression module of the SCID was repeated to reconfirm the diagnosis of MDD, but the entire SCID was not repeated. DSM-5 (American Psychiatric Association, 2013) was not published until after the completion of the study; however, post-hoc review showed that all patients included in the study would have also met criteria for MDD based on DSM-5 criteria. Primary outcomes: depression symptom severity. The Beck Depression Inventory-II (BDI; Beck et al., 1996) was used to assess self-reported depression symptom severity. The BDI is a well-validated self-report measure of depressive  

 

18

symptom severity. It consists of 21 items rated on a 4-point Likert scale and assesses a range of psychological and somatic symptoms of depression. The BDI has been well studied over the past several decades and is sensitive to change and discriminates well between depressed and non-depressed individuals (Beck, Steer, & Garbin, 1988; Richter, Werner, Heerlein, Kraus, & Sauer, 1998). Correlations with clinician-rated measures of depression severity such as the Montgomery-Asberg Depression Rating Scale and the Hamilton Depression Rating Scale range from r = .35 (Schotte, Maes, Cluydts, Doncker, & Cosyns, 1997) to r = .77 (Uher et al., 2008). The BDI was used as the primary outcome measure and was assessed at each study visit. At Weeks 2, 3, and 4, a modified version of the BDI was used that assessed symptoms over the past one week instead of the past two weeks. The Montgomery-Asberg Depression Rating Scale (MADRS; Montgomery & Asberg, 1979) was used as the clinician-rated measure of depression. This 10-item interview measures the following disturbances: sadness, tension, sleep, appetite, concentration, lassitude, numbness, pessimism, and suicidal ideation. The MADRS has demonstrated clinical sensitivity in antidepressant clinical trials and has demonstrated good reliability and concurrent validity with other commonly used clinician-rated measures of depression such as the Hamilton Depression Rating Scale (Davidson, Turnbull, Strickland, Miller, & Graves, 1986; Khan, Khan, Shankles, & Polissar, 2002). The MADRS was administered by independent evaluators blind to treatment condition (CCT or PVT); raters were graduate students. The evaluators were trained on the measure via training with the PI and videotape training. The MADRS was used as a  

 

19

secondary outcome measure and was assessed at baseline, post-treatment, and follow-up visits. Rumination. Rumination was assessed with the Ruminative Response Scale (RRS; Nolen-Hoeksema & Morrow, 1991). The RRS is a 22-item self-report measure assessing the tendency to ruminate in response to sad mood. The scale is rated on a 4point Likert scale; higher scores reflect greater tendency to ruminate. The RRS contains two factor-derived subscales: Brooding and Reflective Pondering (Treynor et al., 2003). Studies of these subscales suggest that brooding is strongly associated with maladaptive emotional outcomes and that reflection may be a more adaptive process (Treynor et al., 2003; Joormann, 2006); however, some research suggests that these two constructs may be less distinguishable in clinically depressed samples (Whitmer & Gotlib, 2011). The RRS has been well-validated in depressed, anxious, and clinically healthy samples (e.g., Nolen-Hoeksema & Morrow, 1991; Nolen-Hoeksema, Morrow, & Fredrickson, 1993; Butler & Nolen-Hoeksema, 1994). The RRS was assessed at baseline, mid-point, posttreatment, and follow-up visits. Environmental reward. The Reward Probability Index (RPI; Carvalho et al., 2011) is a self-report measure designed to assess environmental reward as a way of approximating response-contingent positive reinforcement. The scale consists of 20 items and is rated on a 4-point Likert scale, with higher scores indicating greater levels of environmental reward. The RPI has two subscales: Reward Probability and Environmental Suppressors. The Reward Probability subscale items relate to the number of potential reinforcers and the individual’s ability to engage in instrumental behaviors  

 

20

(Carvalho et al., 2011; for instance “I make friends easily” or “I consider myself to be a person with many skills”). The Environmental Suppressors subscale items relate to the presence of aversive environmental stimuli (Carvalho et al., 2011; for instance, “My behaviors often have negative consequences” or “I have few financial resources, which limits what I can do”). The RPI has been shown to correlate strongly with measures of depression, activity, and pleasurable events (Carvalho et al., 2011). It also demonstrates good test-retest reliability and discriminant validity from social support and somatic anxiety (Carvalho et al., 2011). The RPI was administered at baseline, mid-point, posttreatment, and follow-up visits. Anxiety symptoms. The Beck Anxiety Inventory (BAI; Beck & Steer, 1993) is a commonly used 21-item self-report inventory designed to measure severity of anxiety symptoms. The measure demonstrates high internal consistency and concurrent validity with other assessments of anxiety in psychiatric outpatient populations (e.g., Steer, Ranieri, Beck, & Clark, 1993). The BAI was assessed at baseline, post-treatment, and follow-up visits. Homework compliance. Homework compliance was rated at sessions 2-5 of the behavioral activation treatment. Ratings for Weeks 3 - 5 were made by the therapist and were based on a coding system developed by Busch and colleagues (2010). This system is the first developed specifically to measure homework compliance in BA treatments, and assesses the type of assignment (e.g. single activity, repeated activity), realm of functioning, difficulty level, and extent of completion. Therapists rate the patient’s completion of an activity on a percentage scale (0 – 100% completion) and a categorical  

 

21

scale (3 = fully completed, 2 = partial completion, 1 = made attempt/effort to start, 0 = made no effort to begin assignment). Busch and colleagues (2010) determined that the use of this coding system had a high level of inter-rater agreement and found that homework completion ratings were strongly correlated with treatment change scores (r = .47 and r = .55 respectively for percent completion and categorical completion ratings). Because activity assignments were not assigned before the second session, adherence at Week 2 was rated on a Likert-type scale reflecting the extent to which the activitymonitoring assignment was completed (0 = no completion, 3 = moderate completion, 6 = completed all assignments and brought in written forms). Cognitive Control Tasks NAP Task: Inhibition of emotional processing. The Negative Affective Priming (NAP) task is a computerized negative priming task which assesses inhibition in emotional processing (Joormann, 2004). In the NAP task, participants are asked to respond to a target stimulus in the presence of an irrelevant emotional distractor (i.e., a negative emotional word). On a subsequent trial, the distractor stimulus becomes the target, and the participant’s delay in response latency is considered to reflect the strength of the negative priming effect (with a slower response time reflecting stronger inhibitory control). Participants with higher levels of depressive or ruminative symptoms tend to demonstrate reduced inhibition of negative material. That is, compared to euthymic or low-ruminating individuals, they are quicker to respond to a negatively-valenced target that had been the distractor stimulus in the previous trial. Joormann et al. (2010)

 

 

22

developed the NAP task used in the current study; the task stimuli and procedures are described in detail in Joormann et al. (2010) and are included below for reference. Stimuli. NAP task stimuli consisted of 56 positive, 56 negative, and 16 neutral words. Words were adjectives chosen based on their positive, negative, or neutrally valenced ratings on the Affective Norms for English Words list (Bradley & Lang, 1999). E-prime Psychology Software Tools Inc. version 1.0 software was used to run the task. Procedure and scoring. The NAP task consists of consecutive pairs of trials: a prime trial and a test trial. In each trial, two adjectives are presented simultaneously on the upper and lower halves of the screen, one word in red and one word in blue. Participants are instructed to always ignore the word in red (the distractor) and to respond to the word in blue (the target) by pressing a key to indicate whether the blue word is positive or negative. Participants are not aware of the distinction between a prime or a test trial. Sets of trials are categorized as either “negative priming” trials or “control” trials. During a negative priming trial, the distractor in the prime trial shares the same valence as the target in the subsequent test trial. Therefore, in this condition, the participant’s response to the target in the test trial should be delayed if inhibition of the distractor from the previous prime trial is still activated. During a control condition trial, the distractor in the prime trial does not share the same valence as the target in the test trial, and is a neutral word. Examples of negative priming and control trials are presented in Figures 2 and 3. Participants completed 10 practice trials before completing the full task, which consisted of 5 blocks of 64 prime and test trials (320 trials). The task took approximately  

 

23

15-20 minutes to complete. The words were randomly selected from the word list for each participant, as were the trial sequences within each of the five blocks. The position of the target and distractor words on the screen (upper vs. lower half) was also randomly assigned. In between each trial, a fixation cross appeared on the center of the screen for 500 ms. The words were separated by 2 cm and each letter was approximately 2 cm in size. Reaction times and responses were recorded, and participants were instructed to respond to the target word as quickly and as accurately as possible. Only the responses to the test trials (as opposed to the prime trials) were analyzed. Trials in which the participant responded inaccurately were excluded from analysis, as were trials with outlier response times of < 300 or > 2000 ms (Joormann et al., 2010). Mean response time scores were calculated for four trial conditions: negative priming for negative words, negative priming for positive words, and control trials (for negative and positive words). Additionally, bias scores were calculated for each individual by subtracting the control RT from the negative priming RT (individually for negatively and positively valenced words; Joormann, 2004). Internal Shift Task: Attentional shifting. The Internal Shift Task (IST) is a computerized task that assesses the ability to switch attention between items in working memory in response to emotional and non-emotional facial expressions (De Lissnyder et al., 2012; Demeyer, De Lissnyder, Koster, & De Raedt, 2012). The IST has been shown to have high internal consistency and test-retest reliability (De Lissnyder, Koster, & De Raedt, 2012), and has previously been shown to be associated with worry (Beckwe, Deroost, Koster, De Lissnyder, & De Raedt, 2013), rumination (De Lissnyder, Koster, &  

 

24

De Raedt, 2012), and dysphoric mood (DeMeyer, De Lissnyder, Koster, & De Raedt, 2012) However, to my knowledge, depression treatment effects on IST performance have not been evaluated. IST stimuli and procedures. Stimuli and procedures are identical to De Lissnyder et al. 2012. The 24 neutral and 24 angry face images were taken from the Karolinska Directed Emotional Faces stimuli (Lundqvist, Flykt, & Öhman, 1998), and more specifically, from a validation study of the picture set (Goeleven, De Raedt, Leyman, & Verschure, 2008). E-prime Psychology Software Tools Inc. version 2.0 software was used to run the task. During the IST, participants complete two blocks of emotion and gender condition trials as described above. The order of the two blocks is randomly assigned by the computer program. In each block, participants complete 12 trials in which 10-14 face images were presented on the screen one at a time. Three practice trials precede each block of 12 experimental trials. Participants are instructed to keep a silent mental count of the number of faces they see in each category (angry vs. neutral for the emotional condition, and female vs. male for the gender condition). When a face appears on the screen, they must mentally categorize it (adding it to the existing count of faces) and must press the spacebar as quickly as possible to indicate that they have added that face to the mental count. At the end of each trial, participants report the number of faces from each category as a marker of accuracy. In addition to block condition (emotion vs. gender), trials can be categorized as switch or no-switch trials. A switch trial occurs when the image shown in the trial is of a  

 

25

different category than the preceding image (for instance, a switch from a neutral to an angry face in the emotion condition, or a switch from a female to a male face in the gender condition). A no-switch trial occurs when the image shown in the trial does not change category from the preceding trial (for instance, a neutral face followed by a neutral face in the emotion condition, or a male face followed by a male face in the gender condition). This allows for a calculation of “switch cost”: the difference in average reaction time in a switch trial compared to a non-switch trial (of the same condition). Therefore, IST performance can evaluate general switching capacity (switch cost across both emotion and gender condition), switching capacity for emotional compared to non-emotional material, and switching capacity for specific types of emotional information (angry to neutral, neutral to angry, neutral to neutral, or angry to angry). Median reaction time scores were used for data analyses in order to reduce outlier influence. Both correct and incorrect blocks of items were included in analyses, consistent with previous studies (De Lissnyder, Koster, & De Raedt, 2012). Procedures A table depicting the study measures and procedures at each visit is presented in Table 1. Following the screening visit, eligible participants received baseline assessment, were randomized to treatment condition, and began four weekly sessions of BATD. CCT and PVT exercises were completed in the laboratory immediately prior to each BATD session. Posttreatment assessment took place one week after the fourth and final treatment session (at Week 5). Follow-up assessment took place four weeks following the  

 

26

final treatment session (Week 8). Following final assessment (at the same visit) a booster session of BATD was offered. Participants were encouraged to attend the assessment visits in person but were offered the option of completing assessments by telephone if necessary. The total duration of study participation for each patient was approximately nine weeks. Participants were paid $40 for the completion of the posttreatment and follow-up assessments, for a total of $80. Treatment BATD. Behavioral activation treatment was adapted from the BATD protocol described by Lejuez et al. (2011). This brief protocol has been shown to reduce depressive symptoms in a number of randomized control trials across a range of depressed patient groups, including adult inpatients, individuals in treatment for illicit drug use, and patients with breast cancer (Daughters et al., 2006; Hopko et al., 2003, Hopko et al., 2011). The current study utilized a four-session version of the protocol modified from a previous trial which demonstrated positive effects of a five-session protocol (Magidson et al., 2011). Most efficacy studies of BATD protocols have ranged from 5 to 12 sessions in length (e.g. Magidson et al., 2011; Lejuez et al., 2011). I chose to utilize a four session protocol in order to be consistent with the brief time frame of CCT, which to date has been studied in limited doses within one or two-week periods. The use of this brief protocol is also appropriate for testing the ability of CCT to add benefit to BATD and enhanced the feasibility of the study. An outline of the brief protocol used in the current study is described as follows:

 

 

27

Session 1: Psychoeducation for depression, introduction to treatment rationale and activity monitoring; discussion and identification of patient’s values across different life areas (relationships, occupation/education, recreation, spirituality, etc.) and activities that reflect these values. Session 2: Review activity monitoring and valued activities assignment, begin activity planning. Session 3: Review activity monitoring and planning and generate new plan for the next week. Introduce contracting with friends and family to receive needed support when depressed. Session 4: Review activity monitoring and scheduling and generate new plan for next week. Prepare for end of treatment. Therapists completed a comprehensive four-hour training in the BATD protocol by Dr. Carl Lejuez, followed by weekly supervision from a licensed clinical psychologist with expertise in the protocol throughout the treatment period. Treatment was provided by two doctoral students in psychology. Treatment adherence was monitored with an adherence checklist which therapists self-rated following each session to indicate whether the core strategies and topics were covered. Computerized Experimental Tasks: The active and control versions of the computerized training tasks are described below. Both conditions take approximately 30 minutes to complete and were administered immediately before each of the four BATD sessions.

 

 

28 Cognitive Control Training (CCT) Tasks. CCT consists of two tasks

(completed in the following order): Paced Auditory Serial Addition Task (PASAT; Gronwall, 1977). The PASAT was originally developed to assess cognitive recovery follow brain injury, with a focus on measurement of sustained attention, flexibility, and auditory information processing (Gronwall, 1977). In CCT, a modified computerized version of the PASAT is used (Siegle et al., 2007). Participants are asked to add aurally presented numbers continuously for three five-minute blocks of time. As each new digit is presented, participants must sum it with the digit that was previously presented instead of with the participant’s previous answer. This requires sustained attention to the aurally presented numbers as well as inhibition of encoding of their own responses. Participants are instructed to respond quickly and accurately, and to return to the task as soon as they miss an answer. Task difficulty is equated across participants by adapting the speed based on the patient’s performance. The task starts with a 3000 ms Inter-stimulus Interval (ISI). The speed is increased by 100 ms for every four correct items in a row, and is slowed by 100 ms when a patient misses four consecutive items. Therefore, the average ISI may serve as a marker of task performance. Attention Control Intervention (Wells, 2000). This task was designed to train selective attention to specific information and involves training individuals to attend differentially to multiple auditory sources (e.g., by counting tones, discriminating the location of tones, and moving their attention between auditory sources for a prolonged period). Unlike the PASAT, this task is considered to be “low-load,” and participants  

 

29

must stay focused on the task even when naturally occurring depressive or ruminative thoughts occur (Siegle et al., 2014). The intervention takes approximately 10 minutes to complete. Comparison task. Peripheral Vision Task (C. Moore, personal communication). The PVT task serves as a non-active control condition which does not target the brain regions influenced by the Wells and PASAT tasks. Participants are instructed to focus their eyes on a fixation point in the center of the screen, but to move their peripheral vision around a series of circles on the outer edges of the screen in response to a series of auditory tones. When the sequence of tones ends, they are instructed to indicate the color of the circle that they are now attending to in peripheral vision. The task takes approximately 20 minutes to complete. Power Analysis Effect sizes for CCT have been large in previous investigations of depressed individuals. Compared with a computerized control condition, three doses of stand-alone CCT was associated with an effect size of d = .73 in a community sample of adults with depressed mood (Calkins, McMorran, Siegle, & Otto, 2014). In the current study, each treatment group received an active treatment, which holds the possibility of reducing the between-group differences for the adjunctive interventions. Therefore, a power analysis was conducted anticipating a medium effect size rather than a large one. A priori power analysis using G-Power (version 3.1; Faul, Erdfelder, Lang, & Buchner, 2007) indicated that a sample of 26 patients would provide 80% power to detect a medium effect size (F  

 

30

= .25) using repeated measures ANOVA involving six measurement points (BDI scores assessed at all six study visits; alpha = .05). Statistical Analysis Plan T-tests and Fisher’s exact tests were used to test for baseline differences between the randomized treatment groups, informing whether or not patient characteristics should be controlled for in subsequent outcome analyses. Analyses were planned a-priori using two methods. The first and preferred method was latent growth modeling, a structural equation modeling approach to examining change over time (LGM; also known as latent trajectory modeling or latent curve analysis). LGM allows for modeling individual and between-group differences in trajectories over time and is flexible enough to handle partially missing data and unevenly spaced measurement time points (Curran, Obeidat, & Losardo, 2010; Duncan & Duncan, 2004; Muthen & Curran, 1997). Additionally, LGM lends itself well to evaluation of mediation effects and the effect of baseline scores on change over time (Selig & Preacher, 2009). Despite these strengths, there is a concern regarding the sample size necessary to achieve a model of good fit. Although some have successfully applied LGM to very small samples (for instance, n = 22; Huttenlocher, Haight, Bryk, Seltzer, & Lyons, 1991), it is often preferred that sample sizes of over 100 are used (Curran, Obeidat, & Losardo, 2010). Additionally, there is no clear rule of thumb for estimating the required sample size; according to Muthen & Muthen (2002), this depends largely on aspects of the data that are difficult to estimate in a-priori fashion, such as data distribution, reliability, the extent of missing data, and the strength of the relationships between the variables. Based on these concerns, I planed to first attempt to  

 

31

fit the study data to a latent growth model, and to proceed using LGM for all study analyses if evaluation suggested a good model fit. Tests of model fit included the root mean square error of approximation, Chi Square, and standardized root mean square residual. The second planned method for analysis, in the event that LGM proved inappropriate, was to use repeated measures ANOVA to examine depressive symptoms as functions of the interaction between time and treatment condition. Fisher’s Exact tests would be used to examine differences in the rates of response and remission by treatment group. Secondary analyses were also planned using repeated measures ANOVA to examine between-group differences in rumination (RRS), cognitive control (NAP and IST), environmental reward (RPI), and homework compliance over time. A treatment by time interaction would indicate that CCT vs. PVT differentially affected these potential mediating variables. If these hypothesized effects were obtained, I planned to examine mediation effects using the four-step regression procedures described by Baron and Kenny (1986). Treatment outcome analyses were conducted for the sample of completers (n = 26) and for an intent-to-treat (ITT) sample which included all individuals who were randomized to a treatment condition (n = 34). In the ITT analyses, missing data was accounted for by carrying the last observation forward to all subsequent assessment time points. All tests were conducted using a significance level of p < .05. Effect sizes were calculated using Cohen’s d (interpreted as .2 = small, .5 = medium, .8 = large; Cohen,

 

 

32

1988) and η2 (interpreted as .02 = small, .13 = medium, .26 = large; Murphy & Myors, 2004). RESULTS Sample Characteristics Sample characteristics of participants in the CCT and PVT groups are presented in Table 2. The two groups demonstrated no significant differences in age, gender, education, race, ethnicity, BDI, RPI, RRS, and MADRS scores, or psychiatric medication use. These results remained consistent when including only the smaller sample of treatment completers (n = 26; see Table 3). There were no significant differences on these characteristics between individuals who completed treatment (n = 26) and those who dropped out (n = 8; see Table 4). The mean baseline scores on the BDI (M = 29.6, SD = 10.1) and MADRS (M = 26.6, SD = 7.6) indicate that the overall sample was experiencing moderate to severe levels of depression. Fifty-three percent of the sample had experienced a previous episode of depression. Thirty-five percent (n = 12) of the sample had no comorbid psychiatric disorders; 35% (n = 12) had one comorbid disorder; 21% (n = 7) had two comorbid disorders, and 6% (n = 2) had three comorbid disorders. The comorbid diagnoses were as follows (based on DSM-IV-TR criteria): social phobia (n = 12), eating disorder not otherwise specified (n = 3), generalized anxiety disorder (n = 3), posttraumatic stress disorder (n = 2), dysthymia (n = 2), body dysmorphic disorder (n = 1), obsessive-compulsive disorder (n = 1), trichotillomania (n = 1), somatization disorder (n = 1), panic disorder with agoraphobia (n = 1), agoraphobia without panic disorder (n =  

 

33

1). The number of comorbid diagnoses did not differ between the two treatment conditions (t (31) = -.93, p = .36). Ten participants were taking psychiatric medications at the time of study participation. The medications were as follows: fluoxetine (n = 2), clonazepam (n = 2), amitriptyline (n = 1), atenolol (n = 1), buproprion (n = 1), buspirone (n = 1), citalopram (n = 1), clonidine (n = 1), escitalopram (n = 1), eszopiclone (n = 1), gabapentin (n = 1), lorazepam (n = 1), paroxetine (n = 1), selegiline (n = 1), sertraline (n = 1), trazodone (n = 1), and venlafaxine (n = 1). Treatment Outcome Means and standard deviations for primary and secondary outcome measures are presented in Table 5 (ITT sample) and Table 6 (Completer sample). When BDI scores were entered into a latent growth curve model, indices showed very poor fit (Chi Square = 53.96, Root Mean Square Error of Approximation = .256). Subsequent adjustments to the model based on modification indices and standardized residual values did not improve model fit. I therefore proceeded with treatment outcome analyses using repeated measures ANOVA. Depression symptoms. BDI scores decreased significantly over the course of treatment, as evidenced by a main effect of time on BDI total score in the ITT sample (F (3.29) = 14.01, p = .00, ηp2 = .31) and the completer sample (F (3.30) = 11.91, p = .00, ηp2= .34). Completers demonstrated an average decrease of 11.2 points in BDI score over the course of the study. Concerning Hypothesis 1, there was no main effect of treatment condition (all p values > .70), and the interaction between time and condition on BDI  

 

34

scores was non-significant (ITT sample: F (3.29) = 1.63, p = .18, ηp2= .05); completer sample: F (3.30) = .97, p = .42, ηp2 = .04). The same pattern of results was found when examining MADRS score as a secondary outcome measure of depressive symptoms. MADRS scores decreased significantly over the course of treatment in both the ITT (F (2) = 27.08, p = .00, ηp2 = .46) and completer analyses (F (2) = 23.75, p = .00, ηp2 = .51). No main effect of treatment condition was found (all p values > .50), and the time by condition interaction was non-significant (Total sample: F (2) = .43, p = .66, ηp2 = .01; completers only: F (2) = .36, p = .70, ηp2 = .016). To examine differences in the rates of response and remission by treatment group, recommendations by Riedel and colleagues (2010) were followed: treatment response was defined as a decrease in BDI score of at least 47%, and remission was defined as a BDI score of ≤ 12. When treatment response was examined in the completer sample, rates of remission and response were 34.6% and 38.5% respectively. In the ITT sample, rates of remission and response were 26.5% and 29.4%, respectively. In both the ITT and completer samples, rates of remission and response at Weeks 5 and 8 were higher in the PVT group; however, the difference was not statistically significant. Rates of response and remission by treatment condition are presented in Table 7. Secondary Outcomes Rumination. RRS total scores decreased significantly over time in both the ITT sample (F (2.07) = 14.13, p = .00, ηp2 = .31) and the completer sample (F (3) = 13.45, p = .00, ηp2 = .38). There was no main effect of condition on RRS scores (all p values > .30),  

 

35

and, as a direct test of Hypothesis 2, the interaction between time and condition was not significant (ITT sample: F (2.07) = 1.63, p = .19, ηp2 = .048; completer sample F (3) = .47, p = .71, ηp2 = .02). The brooding subscale of the RRS was examined separately given that it has been shown to be more closely related to depression (Schoofs, Hermans, & Raes, 2010) and cognitive biases (Bernblum & Mor, 2010; Joormann et al., 2010) than the RRS total scale. RRS-brooding scores decreased significantly over time in both the ITT sample (F (2.2) = 7.93, p = .001, ηp2 = .20) and the completer sample (F (2.2) = 6.47, p = .002, ηp2 = .21). There was no main effect of condition (all p values > .85). As the core test relevant to Hypothesis 2, the interaction between time and treatment condition on RRS-brooding was not significant (ITT sample F (2.2) = 2.52, p = .08, ηp2 = .07; completer sample F (2.2) = 1.86, p = .16, ηp2 = .07). Environmental Reward. There was a main effect of time on environmental reward, with RPI-reward probability subscale scores increasing significantly over the course of treatment (ITT sample: F (3) = 14.09, p = .00, ηp2 = .31; completer sample: F (3) = 12.23, p = .00, ηp2 = .36). The effect of condition was not significant (all p values > .48). Relevant to Hypothesis 2, the interaction between time and condition was also not significant (ITT sample F (3) = .91, p = .44, ηp2 = .028; completer sample F (3) = 1.11, p = .35, ηp2 = .048). RPI-environmental suppressor subscale scores also increased significantly over the course of treatment (ITT sample F (2.34) = 6.76, p = .001, ηp2 = .17; completer sample F (3) = 4.81, p = .004, ηp2 = .18). There was no effect of condition on RPI-ES  

 

36

scores (all p values > .28). The interaction between treatment and condition was nonsignificant for both the ITT sample (F (2.34) = 2.27, p = .10, ηp2 = .07) and the completer sample (F (3) = 2.03, p = .12, ηp2 = .08). Anxiety. There was a main effect of time on BAI scores, such that BAI scores decreased over time (ITT sample F (1.61) = 9.08, p = .001, ηp2 = .24; completer sample F (2) = 7.71, p = .001, ηp2 = .27). There was no main effect of condition (all p-values > .70) on BAI scores, and the interaction between treatment condition and time was nonsignificant in both the ITT sample (F (2) = 1.87, p = .16, ηp2 =.06) and the completer sample (F (2) = 1.08, p = .35, ηp2 = .05). Treatment Mediation Although there were no significant differences in outcome by treatment condition, “the absence of differences between two active treatments does not imply that mediated effects are also absent” (Doss et al., 2006). Two equally successful therapies may work through different means; behavioral activation therapy outcomes, for instance, might be mediated by change in environmental reward, while a cognitive therapy might be mediated by change in patients’ cognitions. Therefore, examination of mediation effects may still be informative. In the current study, variables of theoretical interest include cognitive control (IST and NAP task performance), environmental reward (RPI scores), rumination (RRS scores), and homework adherence. Analyses focusing on cognitive control are described in more detail in the “NAP task results” and “IST task results” sections below.

 

 

37 Consistent with Baron and Kenny’s (1986) recommendations, a key requirement

for mediation is to show that the independent variable (in this case, treatment condition) is significantly associated with the hypothesized mediator. In bivariate correlational analyses including only treatment completers, treatment condition (1 = CCT, 2 = PVT) was not significantly associated with the hypothesized mediators – rumination (RRS total r = .02, p = .94; RRS brooding r = .21, p = .30), reward probability (RPI reward probability r = .03 p = .89; RPI environmental suppressors r = .14, p = .49), or homework adherence rating (r = .03, p = .88) - at treatment midpoint. Given that no significant differences were found between treatment conditions on any outcome measure or hypothesized mediator, the two groups were then combined for exploratory analyses that might inform further study of predictors and mechanisms of BATD. Exploratory analyses: What baseline characteristics predict treatment outcome? Hierarchical regression analyses were used to examine the influence of baseline characteristics on follow-up BDI scores alone and in interaction with treatment condition. Baseline BDI scores were entered in the first step of the model, and the potential predictive baseline variable and a dummy variable representing condition were entered in the second step, with the interaction between these two variables entered in the third step. Sex, age, education level, race, psychiatric medication status, and number of comorbid diagnoses were not significantly associated with follow-up BDI scores when controlling for baseline depressive severity (all p values > .40). Furthermore, these baseline variables did not significantly interact with condition to predict BDI outcomes. In  

 

38

addition, when entered alone or in interaction with condition, baseline RRS brooding and reflection scores, and RPI reward probability and environmental suppressor subscale scores were not significantly related to BDI score at follow-up (all p values > .13). Exploratory analyses: Does change in hypothesized mediators predict later change in depression severity? Potential mechanisms of the overall BATD treatment effect could not be tested via statistical mediation due to a lack of no-treatment comparison group. However, it may still be informative to examine changes in proposed mediators and their relationships to change in treatment outcome measures (Doss et al., 2006). If a variable of interest does in fact mediate change in depressive severity, change in the proposed mediator early in treatment should be associated with change in depressive symptom severity later in treatment. Furthermore, it would be expected that the inverse relationship (early change in depressive severity predicting later change in the proposed mediator) would be relatively weaker. Therefore, change scores were calculated for RRS and RPI subscales between baseline and Week 3 (referred to as pre-to-mid RRS and RPI) and were examined as predictors of BDI change scores from Week 3 to 5 (mid-to-post BDI). Next, the reverse analyses were conducted, regressing mid-to-post RRS/RPI scores onto pre-tomid BDI scores. Pre-to-mid RRS change scores were not significantly associated with mid-to-post BDI change scores (b = -.29, t (24) = -1.22, p = .23). However, pre-to-mid BDI change scores were significantly associated with mid-to-post RRS change scores, such that a greater reduction in BDI from baseline to midpoint was associated with a greater  

 

39

reduction in RRS total scores from midpoint to posttreatment (b = -.66, t (23) = -3.095, p = .005). When examining the RRS subscales specifically, a smaller reduction in RRS brooding from baseline to midpoint was associated with an increased reduction in BDI score from midpoint to posttreatment (b = -1.78, t (24) = -2.58, p = .016). However, a larger significant relationship existed between pre-to-mid BDI change scores and mid-topost RRS brooding change scores (b = -.22, t (24) = -3.63, p = .001), suggesting that change in BDI may drive the later change in brooding. Pre-to-mid RRS reflection change scores were not associated with mid-to-post BDI change scores (b = .70, t (24) = .85, p = .40). However, a greater reduction in BDI from baseline to midtreatment was associated with an increase in reflective rumination from midtreatment to posttreatment (b = -.14, t (23) = -2.39, p = .03). A similar pattern of results was found for RPI environmental suppressors: change in this variable from baseline to midpoint did not predict mid-to-post BDI change scores (b = -.54, t (24) = -.98, p = .34); however, a greater reduction in BDI scores from baseline to midpoint did predict an increase in mid-to-post RPI-ES change scores (b = -.20, t (23) = -2.20, p = .038). The relationship between pre-to-mid RPI-RP change scores and midto-post BDI change scores was non-significant (b = .01, t (24) = .025, p = .98), as was the inverse relationship (b = .05, t (23) = .66, p = .52). Finally, homework adherence at midtreatment (computed as the average of the proportion that each assigned activity was completed) was not significantly related with mid-to-post BDI change scores (b = -.01, t (24) = -.098, p = .92). Additionally, pre-to-

 

 

40

mid BDI change scores were not significantly associated with homework adherence at posttreatment (b = .25, t (16) = .36, p = .73). Exploratory Analysis: Does baseline CCT performance predict change in depressive symptoms? A recent study (Siegle et al., 2014) indicated that the benefit of CCT on ruminative symptoms is strongest for individuals who are more engaged in the task at baseline. The pupillary measurement used by Siegle and colleagues (2014) as a marker of engagement was not available in this study; however, the median interstimulus interval time (ISI) from baseline CCT PASAT performance could serve as a measure of initial performance. A lower ISI is indicative of stronger performance. In correlational analyses, stronger baseline performance on the PASAT was predictive of improvement in depressive symptoms at the level of a medium effect (BDI change r = -.36, p =.25 ; MADRS change r = -.51 , p = .09), however, these analyses had a very small sample size (n = 12) and did not reach significance. In contrast, stronger baseline PASAT performance was associated with less improvement in ruminative brooding (r = .32, p = .28). In a more conservative analysis, hierarchical regression models were used to examine whether PASAT performance at baseline predicted follow-up symptom outcomes when covarying for baseline symptom level. PASAT performance was nonsignificant (all p values > .33), and added explanation of 6% variance in depressive symptom outcomes for both BDI and MADRS, and 2% variance in RRS brooding outcomes. NAP Task Results  

 

41 NAP task analyses were performed on the completer sample only. Consistent

with previous studies, participants showed high accuracy rates on the baseline NAP task (control-positive trials: 97%, control-negative trials: 97%, priming-negative trials: 96%, priming-positive trials: 95%). A 2x2 ANOVA was used to examine accuracy rates as a function of condition (priming vs. control) and valence (positive vs. negative). No significant main or interactive effects were found (valence: F (1, 88) = .59, p = .44, ηp2 = .01; condition: F (1, 88) = 2.16, p = .15, ηp2 = .024; valence X condition: F (1, 88) = .002, p = .97, ηp2 = .00). Mean response times for each condition on the NAP task over time are displayed in Table 8. At baseline, a 2x2 ANOVA showed that response times did not differ as a function of NAP condition (F (1, 88) = 1.30, p = .26, ηp2 = .02), valence (F (1, 88) = .024, p = .88, ηp2 = .00), or condition by valence (F (1, 88) = .106, p = .75, ηp2 = .01). The lack of a slowed response in the negative priming trials may indicate a lack of negative priming and therefore a reduced inhibitory control for both negative and positive stimuli. In order to examine the relationship between inhibitory control and depressionrelated variables, NAP scores were calculated for each individual by subtracting response times in the control condition from the response times in the negative priming condition (for positive and negative valences separately). A higher NAP score reflects a greater negative priming effect (and therefore, greater inhibitory control). NAP scores over the course of treatment are presented in Table 8. The NAP-negative score at baseline was not significantly correlated with age (r = -.18, p = .42), sex (r = .14, p = .51), education (r = -.24, p = .27), or baseline BDI (r = .06, p = .80), RRS (r = .27, p = .22), MADRS (r =  

 

42

.20, p = .36), or BAI (r = -.02, p = .93) scores (n = 23 for all correlations). When examining the RRS brooding and reflection subscales separately, NAP-negative scores showed a trend toward a significant correlation with RRS-reflection scores (r = .38, p = .07). RRS brooding was not significantly related to negative priming scores (r = .27, p = .21). Does inhibitory control strengthen as an effect of treatment? If inhibitory control of irrelevant information improved with treatment, response times should specifically increase in the negative priming conditions, and, based on previous research (Joormann et al., 2010), might occur specifically in response to negative stimuli. Therefore, response time as a function of valence and condition were examined at midtreatment, posttreatment, and follow-up. Two-way ANOVAs at each of these time points demonstrated no significant effects for valence, condition, or valence X condition (all p-values > .16, all ηp2 < .026). Therefore, the sample did not exhibit a negative priming effect at any single timepoint in the study. Given that one proposed mechanism of CCT is that it may enhance cognitive control of negative emotions, I also examined whether treatment condition influenced change in NAP scores across treatment. Repeated measures ANOVA of NAP performance at baseline, posttreatment, and follow-up showed no effect of time, condition, or their interaction on NAP negative scores (time: F (2, 36) = .74, p = .49, ηp2 = .039; condition: F (1, 18) = .21, p = .65, ηp2 = .012; time X condition: F (2, 36) = 1.61, p = .21, ηp2 = .082). When examining NAP positive scores, the main effects of time and condition were non-significant (time: F (2, 36) = .58, p = .56, ηp2 = .031; condition: F (1,  

 

43

18) = .087, p = .77, ηp2 = .005). However, there was a significant interaction between time and condition (F (2, 36) = 3.92, p = .029, ηp2 = .179; see Figure 4). As depicted in Figure 4, for those assigned to CCT, NAP positive scores decreased between baseline and posttreatment and increased again between posttreatment and follow-up. For those assigned to PVT, NAP positive scores increased between baseline and posttreatment and then remained relatively consistent between posttreatment and follow-up. Is change in inhibitory control associated with symptom change? To reflect the magnitude of change in inhibitory control over time, NAP change scores (for negative and positive stimuli separately) were calculated by subtracting the priming scores at baseline from the priming scores at posttreatment. A positive value for this variable indicates increased priming (i.e., improvement in inhibitory control) over time. Bivariate correlations were then used to examine whether change in inhibitory control was associated with symptom change. NAP change scores were not associated with BDI (r = .21, p = .37), MADRS (r = -.15, p = .52), RRS brooding (r = .12, p = .61), or RRS reflection (r = .10, p = .69) change scores. To examine this question differently, I also investigated whether individuals who responded to treatment demonstrated greater improvement in inhibitory control. Individuals classified as treatment responders at follow-up did not differ significantly from non-responders in NAP change scores (NAP-negative: t (20) = .19, p = .58, d = .24; NAP-positive: t (20) = -.31, p = .76, d= .13). Does baseline inhibitory control predict symptom change? In exploratory analyses, we investigated whether NAP performance at baseline was predictive of  

 

44

treatment outcome or related characteristics alone or in interaction with treatment condition. Symptom measures at follow-up (BDI, MADRS, RRS, RPI) were regressed onto baseline NAP negative scores, treatment condition, and their interaction term. The baseline score for the symptom measure of interest was included as a covariate. Results are displayed in Table 9. NAP negative scores at baseline, when alone and in interaction with treatment condition, did not significantly predict BDI at follow-up. However, there was a trend toward a significant interaction between treatment condition and NAP negative scores in predicting MADRS scores at follow-up (b = -.033, t (18) = -1.86, p = .08, R2 change = .088). Examination of this interaction effect revealed that in the CCT group, baseline NAP negative scores were trending toward a significant negative association with MADRS scores at follow-up (b = -.024, t (9) = -2.19, p = .06, R2 change = .19) such that individuals with increased inhibitory control of negative emotional material at baseline demonstrated lower follow-up MADRS scores. However, in the PVT group, NAP negative scores were unrelated to follow-up MADRS scores (b = .035, t (8) = .83, p = .43, R2 change = .04). Conversely, NAP negative scores at baseline were significantly associated with increased RRS brooding scores at follow-up when covarying baseline brooding (b = .01, t (19) = 2.28, p = .04), but did not significantly interact with treatment condition (p = .12) in this model. Finally, NAP negative scores at baseline, alone and in interaction with treatment condition, did not significantly predict RRS-Reflection scores at follow-up. (all p values > .18)  

 

45

IST Results IST analyses were performed on the completer sample only. The means and standard deviations for IST trials by condition and switch type are presented in Table 10. As expected, across all conditions, participants responded more slowly to switch trials than to non-switch trials (all p values < .002). A 2x2 ANOVA of reaction times by Condition (emotion, gender) and Switch Type (switch, no-switch) showed main effects of both condition and switch type (condition: F (1, 98) = 4.26, p = .042, ηp2 = .04; switch type: F (1, 98) = 18.70, p = .00, ηp2 = .16). Specifically, participants responded more quickly to gender trials than to emotion trials, and to no-switch trials than to switch trials. However, the Condition by Switch type interaction was not significant (F (1, 98) = .044, p = .83, ηp2 = .00). Additionally, there was no significant difference between gender switch cost and emotion switch cost at baseline (t (22) = 1.15, p = .26; see Table 10 for all values). Emotion and gender switch costs were highly correlated (r = .61, p = .002). I also examined the effect of emotional valence within the emotion condition. A 2x2 ANOVA on reaction times with Valence (angry, neutral) and Switch type (switch, no switch) showed a main effect of switch type (F (1, 96) = 9.31, p = .003, ηp2 = .088). Participants responded more slowly to switch trials (either an angry to a neutral face or a neutral to an angry face) than to non-switch trials. There was no main effect of valence (F (1, 96) = .01, p = .92, ηp2 = .00), and the interaction between valence and switch type was non-significant (F (1, 96) = .00, p = .96, ηp2 = .00). Emotion switch cost at baseline was not significantly correlated with age (r = .27, p = .20), baseline BDI (r = -.06, p = .78), MADRS (r = -.14, p = .51), BAI (r = -.01,  

 

46

p = .96), RRS total (r = .02, p = .92), RRS brooding (r = -.22, p = .29), or RRS reflection scores (r = .02, p = .92) (n = 24 for all correlations). There was no significant difference in emotion switch cost between men and women (F (1, 22) = 1.03, p = .32), although Cohen’s d (d = .43) indicated a moderate effect size, with women demonstrating a larger emotion switch cost than men. There were no significant differences in emotion switch cost by race (F (2, 21) = .55, p = .59, ηp2 = .05) or education level (F (3, 20) = .87, p = .47, ηp2 = .12). Does switching capacity strengthen as an effect of treatment? Given that CCT may differentially enhance attentional switching capacity compared to PVT, I examined whether treatment condition influenced change in switch cost scores over the course of treatment. Repeated measures ANOVA of switch cost scores at baseline, posttreatment, and follow-up showed no significant effect of time, treatment condition, or their interaction on emotion switch cost (time: F (2, 30) = 2.26, p = .12, ηp2 = .13; condition: F (1, 15) = .06, p = .81, ηp2 = .00; time by condition: F (2, 30) = 1.52, p = .24, ηp2 = .092). When examining gender switch cost scores, there was a main effect of time (F (2, 32) = 4.46, p = .02, ηp2 = .22), showing that gender switch costs lessened over time. However, there was no main effect of condition (F (1, 16) = .38, p = .55, ηp2 = .02) or time by condition (F (2, 32) = .95, p = .40, ηp2 = .06). Because a number of participants did not have complete IST data from all three time points, the sample size was limited in this analysis (for emotion switch costs, nCCT = 7, nPVT = 10; for gender switch costs, nCCT = 7, nPVT = 11).

 

 

47 I also examined whether the main effects of Switch Type and/or Emotion

Condition found at baseline remained present through posttreatment and follow up. At posttreatment, a 2x2 ANOVA showed a main effect of switch type (F (1, 66) = 12.92, p = .00, ηp2 = .16) but no longer showed a main effect of emotion condition (F (1, 66) = 2.03, p = .16, ηp2 = .03). The switch type by emotion interaction was not significant (F (1, 66) = .052, p = .83, ηp2 = .001). At follow-up, main effects of both switching condition (F (1, 76) = 19.31, p = .00, ηp2 = .20) and emotion condition (F (1, 76) = 4.98, p = .029, ηp2 = .063) were found, such that participants responded more quickly to non-switch trials compared to switch trials, and to gender trials compared to emotion trials. The condition by switch interaction was not significant at follow-up (F (1, 76) = .10, p = .76, ηp2 = .00). Is change in switching capacity associated with symptom change? To reflect the magnitude of change in switch costs over time, switch cost change scores (for emotion and gender conditions separately) were calculated by subtracting the switch costs at posttreatment from the switch costs at baseline. A positive value for this variable indicates a larger decrease in switch cost over time. Bivariate correlations were then used to examine whether change in switch costs were associated with symptom change. Emotion switch cost change scores were not significantly associated with BDI (r = .23, p = .37), MADRS (r = .22, p = .40), or RRS reflection (r = -.38, p = .13) change scores. There was a trend toward a negative correlation between emotion switch cost change and RRS-brooding change from baseline to posttreatment (r = -.43, p = .086) To examine this question differently, I also investigated whether individuals who responded to treatment demonstrated greater change in switch costs over time.  

 

48

Individuals classified as treatment responders showed a greater average reduction in emotion and gender switch costs at follow-up, but these differences were not significant (emotion switch cost: t (18) = -.69, p = .50, d = .33; gender switch cost: t (18) = -1.18, p = .25, d= .56). Does switching capacity at baseline predict later symptom change? In exploratory analyses, we investigated whether IST performance at baseline was predictive of treatment outcomes. Symptom measure scores at follow-up were regressed onto baseline Emotion Switch Cost scores and Gender Switch Cost scores (separately) alone and in interaction with treatment condition. The baseline score for the symptom measure of interest was included as a covariate in the model. Results are displayed in Table 11. Baseline emotion switch cost, treatment condition, and their interaction did not predict BDI, MADRS, RRS-brooding, or RRS-reflection scores at follow-up. Additionally, baseline sex switch cost, condition, and their interaction did not predict BDI, MADRS, RRS-Brooding, or RRS-Reflection scores at follow-up. DISCUSSION The current study tested whether CCT, a neurocognitive intervention previously shown to reduce depression symptoms, enhanced the effects of BATD when compared to a control condition (PVT). I hypothesized that individuals receiving BATD + CCT would demonstrate a significantly greater reduction in depressive symptoms than individuals who received BATD + PVT. Results showed that both treatment groups demonstrated a significant decline in depression symptom severity, ruminative brooding and anxiety, as well as an improvement in environmental reward. However, individuals  

 

49

who received CCT did not differ significantly from the PVT group on any of these clinical outcome measures. This finding poses a contrast to two studies in which adjunctive CCT was shown to be clinically beneficial. In the first test of CCT (Siegle et al., 2007), the addition of six sessions of CCT to an intensive outpatient program for severely depressed patients led to significant clinical improvement compared to treatment as usual. A more recent study showed that the addition of CCT to five sessions of tDCS for depression was associated with more sustained symptom reduction than tDCS alone (Segrave et al., 2013). Considering secondary aims, I hypothesized that relative to the BATD + PVT condition, the BATD + CCT condition would be associated with greater improvement in processes—ruminative brooding and cognitive control—thought to be specifically sensitive to CCT. Specifically, I hypothesized that switching performance (as measured by the IST task) would improve as indicated by increased response latency, and that inhibitory control of negative material (as measured by the NAP task) would improve as indicated by an increased priming effect toward negative stimuli. Finally, I hypothesized that ruminative brooding would decrease in the BATD + CCT condition relative to the BATD + PVT condition. Study results were not supportive of these hypotheses. Changes in brooding, switching performance, and inhibitory control of negative material did not differ significantly by treatment condition. In addition, although the patterns of change in NAP positive scores did differ between treatment groups, this effect did not occur in the hypothesized direction and was found only for positive emotional stimuli. Specifically, for those assigned to CCT, inhibitory control of positive stimuli decreased  

 

50

during the treatment period and increased during the follow-up period. For those assigned to PVT, inhibitory control of positive stimuli increased during the treatment period and then remained relatively consistent during the follow-up period. With regards to Aim Two, I also hypothesized that changes in ruminative brooding and cognitive control would mediate the relationship between treatment condition and treatment outcome and the relationship between treatment condition and homework adherence. However, this mediation effect was not present as evidenced by the lack of relationship between treatment condition and the proposed mediators. This finding diverges from the recent work of Siegle and colleagues (2014), who found that, compared with treatment as usual, adjunctive CCT led to unique reductions in ruminative brooding. Several factors may account for the failure of CCT to augment BATD in the current study. Although the total overall dose of CCT provided in the current study was consistent with previous studies (e.g., Calkins et al., 2013), the timing of CCT sessions differed. I administered CCT sessions immediately before four weekly therapy sessions, a schedule that was consistent with the type of dosing that might feasibly be offered as an adjunct to outpatient care. Previous studies have administered CCT sessions either 2-3 times per week (Siegle et al., 2007, Calkins et al., 2013), or daily for 1-2 weeks (Segrave et al., 2013; Brunoni et al., 2014). Thus, one possible explanation for the failure of CCT in the current study may be that the spacing between sessions was not sufficient to influence depressive symptoms. The current study indicates that CCT in a weekly format

 

 

51

does not add benefit to BATD. Future studies could consider supplementing outpatient treatment with daily at-home CCT to investigate the influence of an increased dose. In addition, the current study is the first to my knowledge to pair CCT with a BA treatment. BA is a well-supported therapy that has demonstrated very strong effects, comparable to antidepressant medication and cognitive therapy (Dimidjian et al., 2006), when administered as a full-length treatment package (typically 12-20 sessions). Clinical gains have been relatively attenuated when BA has been offered in brief formats, and because it does not contain a cognitive component, BA provided a suitable platform for CCT to supplement. However, it is possible that even in a brief format, the effects of BATD were robust enough to overshadow the benefits of CCT. Indeed, the active comparison condition in this study (BATD + PVT) yielded an average reduction of 13 points on the BDI, with thirty-eight percent of the ITT sample meeting criteria for remission. In contrast, in the study by Siegle and colleagues (2007) that provided initial support for CCT, participants within the active comparison condition (an intensive outpatient program) saw no change, or possibly even an increase in BDI score over time. In addition to the strong effects on depressive symptoms across both conditions, treatment was associated with reductions in ruminative brooding and IST gender switch costs. Based on the theory underlying CCT, I had hypothesized that these cognitive variables would show more improvement over time in the CCT condition. The fact that both treatment conditions led to changes in these proposed mediator variables suggests a possible shared mechanism in CCT and BA treatment. The lack of mechanisms unique to CCT found here is in contrast to the work of Siegle and colleagues (2007), who found  

 

52

that adjunctive CCT was associated with increased activation of frontal brain regions. However, in Siegle’s (2007) study, the treatment as usual group did not undergo parallel neuroimaging analyses. Therefore, it may be that the changes found in prefrontal activation represent a non-specific consequence of improvement from depression rather than a unique effect of CCT. This explanation would be in line with several neuroimaging studies demonstrating enhanced frontal activation following successful treatment for depression using antidepressant medications (Fales et al., 2009; Mayberg et al., 2000) and cognitive behavior therapy (Ritchey, Dolcos, Eddington, Strauman, & Cabeza, 2011). Despite the lack of evidence of unique CCT treatment effect within this study, the possibility remains that CCT may be beneficial for a select group of patients with depression. MDD is a heterogeneous disorder (Winokur, 1997), and CCT targets a very specific mechanism (executive control) proposed to play a role in the disorder. It may be that individuals will benefit from CCT only to the extent that they demonstrate deficits in this area (Siegle et al., 2014). It may also be the case that individuals must participate in CCT in a specific manner in order to receive benefit; as Siegle and colleagues (2014, p. 456) suggest, “some critical expenditure of cognitive resources” may be necessary in order for CCT to act on the cognitive substrates that might lead to emotional change. This is supported by an expanded analysis of the data from the initial trial of CCT, in which Siegle and colleagues (2014) examine predictors of response to CCT. With a new analysis strategy, Siegle et al. found that while nearly all participants (assigned to treatment as usual (TAU) or TAU + CCT) saw improvement in depression symptoms, the  

 

53

CCT group demonstrated significantly fewer ruminative brooding symptoms. Importantly, the authors also found that higher scores on a pre-treatment physiological measure of task focus predicted reduction in brooding in the CCT group, but not in the TAU group. Their finding indicates that CCT may be most valuable for individuals who were strongly engaged with the task at the initiation of treatment. The current study did not contain a measure of task engagement similar to that used by Siegle and colleagues (2014). CCT performance at baseline, perhaps the index most similar to “task engagement”, was moderately correlated with BDI and MADRS follow-up outcomes, a finding consistent with Siegle’s (2014) work. However, baseline CCT performance did not add unique prediction of outcomes when covarying for baseline symptom levels, and the sample size of this analysis was small (n = 12). I also examined whether CCT was differentially effective for individuals with reduced cognitive control of emotional material, as operationalized by performance on the IST and NAP tasks. Performance on the IST (a measure of switching performance) did not significantly interact with treatment condition to predict outcome. In the NAP task analyses, a trend was found toward an interaction between baseline inhibitory control of negative emotional stimuli (NAP negative scores) and treatment condition in the prediction of follow-up MADRS scores. That is, in the CCT group, but not the PVT group, individuals with higher inhibitory control of negative stimuli at baseline demonstrated reduced MADRS scores at follow-up. This finding is consistent with the hypothesis that CCT effects are moderated by the baseline availability of cognitive resources. However, it should be interpreted with caution due to the small sample size (n  

 

54

= 23) and the lack of similar pattern of moderation when examining BDI or RRS scores as measures of treatment outcome. Although it is suggested that CCT may act on depression through improvements in cognitive control and rumination, the current study does not provide support for these proposed mechanisms of action. When compared to BA plus a well-designed sham condition, BA+ CCT did not lead to significantly greater reductions in depression or rumination, or to improvements in cognitive control. One possible interpretation of these results is that CCT’s effects may be due to placebo or expectancy effect. The designs of previous studies largely have been unable to rule out this possibility because they have not compared CCT to a sham condition. In one of the few studies to use a sham comparison condition rather than TAU or no-treatment comparison, Calkins et al. (2014) did find positive effect of CCT on depressive symptoms compared to PVT. However, a later study (Moshier, Molokotos, Stein, & Otto, in press) with an identical design (3 sessions of CCT over 2 weeks in a sample of participants with elevated BDI scores) showed no effect of CCT compared to PVT. The only other study to examine CCT relative to a sham condition did so in adjunct to tDCS. Segrave and colleagues (2014) found that CCT plus tDCS, when compared to sham CCT plus tDCS and CCT plus sham tDCS was associated with sustained reduction in depressive symptoms. The results suggest that tDCS and CCT in combination are acting as more than placebo, but their combination does not allow us to understand the unique influence of CCT. Given this pattern of findings, future research needs to be designed to evaluate whether CCT works through its hypothesized mechanisms rather than through expectancy effect.  

 

55

This will require the use of adequate comparison conditions that offer similar expectancy effects, such as sham computerized tasks. Furthermore, we should continue to examine mediators of CCT effects, with a focus on identifying those that are unique to CCT compared to active treatments or placebos. This has been a challenge across psychosocial and pharmacological treatments for depression, with some arguing that a substantial proportion of the effects of treatment may be due to placebo effect and/or natural changes in course of depression (Kirsch & Sapirstein, 1998). Consistent with this, rates of response to pill placebo for depression have been shown to average 30% (Walsh, Seidman, Sysko, & Gould, 2002). Neuroimaging studies may be one way to distinguish the theorized effects of treatment from placebo effects. Studies have identified changes in brain regions unique to specific treatments, such as fluoxetine or CBT, and not seen in placebo (see Benedetti, Mayberg, Wager, Stohler, & Zubieta, 2005). However, one potential problem for research on the neural mechanisms of CCT is that the placebo effect in depression has been shown to be associated with increases in DLPFC activation, the very area which CCT targets. Thus, the brain regions that might change if CCT works the way it has been hypothesized to might also be expected to change in response to placebo, making it difficult to distinguish between the two. Brief Behavioral Activation Therapy for Depression The present study, although focused on the clinical efficacy of adjunctive CCT, also provides support for the further study and use of BATD for individuals with MDD. Following four sessions of BATD, patients demonstrated an average 11-point reduction in BDI score and 8-point reduction in MADRS score. Thirty-five percent of the sample  

 

56

reached remission status by the follow-up visit. These results are consistent with a growing body of evidence supporting the use of brief BA treatments. The majority of these studies have examined time-limited BA protocols in individuals with elevated, but not necessarily clinical levels of depression (e.g., Magidson et al., 2011; Daughters et al., 2008, Gawrasiak & Hopko, 2009). Our results expand upon this work to show that brief treatment may be useful for those with a clinical-level diagnosis of MDD. They replicate the results of an open trial conducted by Gross and Haren (2011) which demonstrated that four sessions of BA in a primary-care setting significantly reduced MDD symptoms in veterans. An important next step will be to compare time-limited BATD to a control condition within a sample of clinically depressed patients. The expanding support for brief BA treatments has important implications for depression treatment efforts. Estimates suggest that nearly 40% of patients who experience a major depressive episode choose not to seek treatment, and that cost is often a major factor in this decision (Mojtabai, 2009). Furthermore, fewer than half of individuals who do seek treatment for depression receive adequate care (Chermack et al., 2008; Prins et al., 2010; Witt et al., 2011). BA has potential to help alleviate the large unmet need for affordable, accessible, evidence-based treatment for depression. Research has shown that non-specialists can be trained to effectively administer BA (Ekers et al., 2011), which would allow for an increased number of available clinicians as well as a lower cost of intervention. BA can also be flexibly adapted for different settings and populations; for instance, programs have been modified to focus on smoking cessation, substance use disorders, and HIV medication adherence in the context of  

 

57

depressive symptoms (Magidson et al., 2011; Daughters et al, 2008; Macpherson et al., 2010). Finally, the format of BA also lends itself to be used in low-cost and transportable modes of intervention, such as bibliotherapy and Internet-based treatment. Recent studies have provided initial support for the effectiveness of these delivery methods (O’Mahen et al., 2013; Carlbring et al., 2013). Given its brevity, BA may be an important intervention in a stepped-care approach which offers different levels of interventions. From this perspective, it would be particularly useful to understand which patients with MDD are likely to benefit from brief BA treatment and which patients need more resource-intensive treatment. The current study did not clarify this question, finding that baseline clinical measures (rumination, environmental reward) and characteristics such as sex, age, education level, race, psychiatric medication status, and psychiatric comorbidity were not significantly associated with treatment outcome after controlling for baseline severity of depressive symptoms. This may reflect a problem seen across the literature - that variables that have been shown to predict outcome or course of illness alone, such as cognitive style, lose significance when considered in addition to baseline depression symptoms (Otto et al., 2007). Continued research with larger samples is needed so that we may begin to identify predictors of treatment outcome that will allow for more tailored treatment recommendations. The current study also examined potential mechanisms of BATD effects, with focus on changes in ruminative brooding, environmental reward, and adherence to scheduled activities. Because of the lack of comparison group in the current study, formal  

 

58

mediation analyses of BATD effects were not possible. Instead, I examined midtreatment changes in proposed mediators of BATD (brooding, environmental reward, and homework adherence) and their relationships to subsequent change in depressive symptoms. If these variables were mediators of outcome, one would expect that early changes in these variables would predict later changes in depression symptom severity. However, the opposite pattern was found: early changes in depressive symptoms (as measured by the BDI) were more strongly related with later changes in rumination and environmental reward. In addition, the relationship between homework adherence and later change in BDI was non-significant. The lack of mediation could indicate that BA exerts change in depression through other mechanisms, such as cognitive change. Alternatively, it may be that the measures used, such as the RPI and homework adherence ratings, do not accurately capture changes in activity level or positive reinforcement (theorized mechanisms of BA). For instance, although homework adherence ratings assign equal weight to every activity that a patient completes, it is likely that each assigned activity in BA is differentially reinforcing for a patient. Furthermore, such ratings of activity completion may not capture other ways that patients initiate BA strategies in their daily lives. Future research on the mechanisms of change in BA might benefit from more precise measurement strategies such as daily activity diaries or ecological momentary analysis. Cognitive Control In the current study, I examined two measures of cognitive control of emotional material as potential mediators of CCT treatment effects. Both measures, the Negative  

 

59

Affective Priming (NAP) Task and the Internal Shift Task (IST), have been linked with rumination and depressive symptoms in previous work. As noted above, adjunctive CCT was generally not associated with differential clinical change between groups. However, exploration of data from these tasks across the treatment groups has potential to inform research into cognitive control processes in patients with depression more generally. Inhibitory control – NAP task. Previous research on the NAP task has compared clinical and non-clinical samples, showing that patients with MDD demonstrate reduced inhibitory control of negative but not positive emotional information, while healthy controls show priming effects (indicating inhibitory control) to both negative and positive information (Joormann & Gotlib, 2010). This difference could not be replicated in the current study because there was no healthy comparison group. However, within-subject tests of baseline NAP performance revealed no significant main effects of priming condition or valence of stimuli. Thus, in contrast to previous studies, participants in the current study did not demonstrate a priming effect for negative or positive stimuli. These results may indicate reduced cognitive control in response to emotional content more globally (as opposed to only negative stimuli). Although it is unclear why these results differ from Joormann & Gotlib in this way, they are consistent the work of De Lissnyder and colleagues (2012), who found that although dysphoric individuals demonstrated deficits specific to negative emotional content, clinically depressed individuals demonstrated a broader, more global cognitive impairment in set-shifting ability. One possibility is that in more severely depressed individuals, exposure to negative emotional stimuli has a more sustained influence on  

 

60

inhibitory control; thus they may be less likely to be influenced by the valence of the stimuli appearing on a trial-by-trial basis on these types of experimental tasks (De Lissnyder et al., 2012). In order to better understand the clinical relevance of inhibitory control processes, I examined the relationships between clinical characteristics and NAP task performance at baseline, and also analyzed changes in NAP task performance over the course of treatment. Consistent with previous work in a sample including healthy controls and both currently and formerly depressed individuals (Joormann & Gotlib, 2010), I found a trendlevel positive association, reflecting a medium effect size (r = .38), between inhibitory control of negative information (NAP negative scores) and baseline scores on the reflection subscale of the RRS. However, in the current study, correlations between baseline NAP performance and all other clinical measures were small in size and nonsignificant. This diverges with the results of Joormann and Gotlib (2010), who found that in depressed patients, increased levels of brooding were strongly (r = -.41) related to reduced inhibitory control of negative information. Joormann and Gotlib (2010) also found that unlike depressed patients, formerly depressed individuals did not demonstrate a reduced inhibitory control of negative information. Interestingly, they found that these remitted depressed individuals showed a failure to inhibit positive information, possibly demonstrating a positive attentional bias. This would suggest that with successful treatment, depressed patients might gain inhibitory control for negative material and begin to show a bias toward positive information. In the current study, I found no evidence of change in inhibitory control of  

 

61

negative information across the course of treatment. However, participants who received CCT showed a reduction in NAP positive scores between baseline and posttreatment, and a subsequent increase during the follow-up period. Although this pattern may be a sign that CCT leads to a positive attentional bias (at least acutely), this result must be interpreted with caution, as the change in NAP positive score was not associated with any measure of clinical outcome. Furthermore, the decrease in NAP positive scores reflects a reduction in inhibitory control of positive information, which is at odds with the hypothesis that CCT would increase inhibitory control. More research is needed to understand how the NAP task relates to clinical processes and to assess changes in NAP performance in response to treatment. It may be that individuals in the current study did not improve enough clinically to see a change in inhibitory control as was apparent in the sample of formerly depressed patients (Joormann & Gotlib, 2010). It is also possible that changes in inhibitory control may follow symptom change, and that eight weeks was too short of a period in which to detect this change. In exploratory analyses, I also examined whether baseline inhibitory control of negative material predicted treatment outcomes alone or in interaction with treatment condition. Findings showed that although baseline inhibitory control of negative material was unrelated to depression symptom severity at follow-up, it did significantly predict higher brooding scores at follow-up. There was no significant treatment condition by NAP score interaction, suggesting that this effect reflected changes related to BA treatment rather than CCT. These results suggest that BA may be more likely to act on  

 

62

depressive brooding in individuals with lower levels of pre-treatment inhibitory control of negative emotional material. Given that brooding has been uniquely associated with negative outcomes such as suicide ideation (Miranda & Nolen-Hoeksema, 2007) and relapse following depression treatment (Michalak, Holz, & Teismann, 2011), there is clinical value in reducing depressive brooding in addition to depression symptoms. In the future, it may be worth examining whether inhibitory control moderates reduction in brooding across other types of treatments in addition to BA. It may also be important to understand why individuals with higher inhibitory control of negative material may not experience reduction in brooding through BA treatment and how BA could be modified in order to do so. However, further replication of this finding is needed given the exploratory nature of the analysis and the limited sample size. In sum, the current examination of NAP task performance at baseline and over the course of treatment found that depressed participants failed to inhibit both positive and negative emotional information and suggested that inhibition of negative material did not change over the course of treatment. Furthermore, NAP task performance was not significantly associated with baseline brooding (r = .27). This finding is inconsistent with the only previous study to examine NAP task performance in a clinically depressed sample, which found a moderate and negative correlation between NAP negative scores and brooding (r = -.41; Joormann et al., 2010). Because both of these studies have been limited by small sample sizes, more work is needed in larger samples to determine the clinical relevance of inhibitory control within depressed populations. Understanding how variables other than clinical severity of depression and rumination might relate to  

 

63

inhibitory control may help to advance the research in this area. For instance, Joormann’s study found evidence that improved inhibitory control was associated with the use of adaptive emotion regulation strategies such as reappraisal (Joormann & Gotlib, 2010). Future work might continue along this line by examining how inhibitory control relates to other processes involved in depression such as reward sensitivity or autobiographical memory. Moreover, little is known about whether the NAP task is sensitive to factors such as current affective state, working memory load, fatigue, or effort. Experimental studies which measure or manipulate these factors may lead to a more comprehensive understanding of the NAP task and the role of inhibitory control in depression. Switching performance – IST task. I also assessed cognitive control with the Internal Shift Task (IST), which assesses the ability to switch attention between items in working memory in response to images of faces. The IST allows for assessment of response time performance when emotional expression is task-relevant (the emotion condition) or not task relevant (gender condition), and under conditions of switch trials (the trial stimuli presented is of a different category than that of the previous trial) or noswitch trials (the trial presents stimuli within the same category as the previous trial). At baseline, participants demonstrated a slower response to switch trials (compared to non-switch trials) and a slower response to trials in which emotional expression was task-relevant (compared to trials where gender was the task-relevant feature). The slowed response to switch trials across both gender and emotion conditions is consistent with findings by De Lissnyder et al. (2012), who found that clinically  

 

64

depressed individuals demonstrated a general switching impairment. That is, unlike dysphoric individuals, who demonstrated a switching impairment specifically in the emotion condition, depressed individuals demonstrated a switching impairment in both conditions. Although the lack of control group in the current study prevents from making the conclusion that switching performance was impaired relative to healthy individuals, the results are broadly consistent with this previous work. De Lissnyder and colleagues (2012) also found that IST switch costs (the average difference between response times to switch and non-switch trials) were positively correlated with rumination and depression symptoms in depressed individuals. However, in the current study I found no significant relationship between baseline IST performance and clinical measures. In addition, baseline switch costs did not predict treatment outcome when controlling for baseline levels of symptom severity. In exploratory analyses, I also examined change in switch costs as a function of time, treatment condition, and their interaction. Results indicated that over time, participants demonstrated increased ability to switch attention in the gender condition of the IST; however, no change occurred in performance on the emotion condition. This is the second study to demonstrate enhanced performance on the gender condition of the IST over the course of time (Onraedt & Koster, 2014). This improvement may simply be a practice effect from repeated exposure to the IST task. However, the current data suggest that it is more than this. Compared to non-responders, treatment responders demonstrated a trend toward a greater reduction in gender switch cost over time. Thus, change in IST performance in depressed individuals may be related to clinical gains over  

 

65

the course of treatment. If this finding can be replicated in larger samples of patients receiving treatment for depression, it will then be important to investigate whether change in switching performance is a mechanism or consequence of clinical symptom change. Limitations Findings of this study must be interpreted in the context of a number of limitations. Because both treatment groups received BATD and there was not a notreatment comparison condition, I cannot confirm that the reduction in depressive symptoms was due to BATD rather than another unknown cause. Second, the small sample size may have prevented detection of predictors or mediators of treatment response. Baseline clinical and demographic measures did not significantly predict treatment outcome and effect sizes were small. The most promising predictor of outcome was baseline performance on the PASAT task, with stronger performance predicting greater improvement in BDI and MADRS scores at the level of a medium but nonsignificant effect. However, sample size was very limited for this analysis, and was also limited for detecting changes over time in the cognitive control tasks as a function of treatment condition. Additionally, the study does not allow conclusions to be made regarding the effects of CCT or BATD past the one-month follow-up period. The sample was predominantly White and highly educated, with more than half the sample having obtained a Bachelor’s degree or higher. Therefore, caution should be taken in generalizing these results to non-White or less educated patient groups. Finally, although participant report suggests that drop-out did not occur due to unacceptability of CCT, a  

 

66

significantly greater number of participants in the CCT condition dropped out of the study compared to the PVT condition. Future research should address these limitations using larger, more diverse samples, collecting information about CCT acceptability, and by increasing the length of the follow-up period following treatment. CONCLUSION The current project represents an assessment of a mechanism-focused, low-cost augmentation strategy for the treatment of depression. Weekly CCT was not found to add clinical benefit to a four-session BATD treatment, and CCT augmentation was not associated with relative increases in cognitive control, ruminative brooding, or homework adherence. These results suggest that the effects of CCT may not be as robust as previous studies have suggested. This is consistent with a recent re-analysis of the most comprehensive CCT study in clinically depressed patients (Siegle et al., 2014) suggesting that CCT may only benefit those who are able to allocate sufficient attentional resources toward the task. Despite the failure of CCT to aid BATD when administered in weekly format, there is continued need to investigate CCT. Future work should prioritize understanding for whom CCT may be effective and the conditions that optimize its effectiveness.

 

 

67 TABLES

  Table 1. Study measures and procedures by visit. Baseline (Week 1)

Week 2

Week 3

Week 4

Endpoint (Week 5)

Followup (Week 8)

BDI

X

X

X

X

X

X

MADRS

X

X

X

BAI

X

X

X

RPI

X

X

X

X

RRS

X

X

X

X

NAP task

X

X

X

X

IST task

X

X

X

X

Screening (Week 0) Informed consent

X

SCID-IV

X

Personal Data form

X

Homework adherence

X

X

X

CCT or PVT

X

X

X

X

BATD session

X

X

X

X

X

Note. SCID-IV = Structured Clinical Interview for DSM-IV; BDI = Beck Depression InventoryII, MADRS = Montgomery-Asberg Depression Rating Scale; BAI = Beck Anxiety Inventory; RPI = Reward Probability Index; RRS = Ruminative Response Scale; NAP task = Negative Affective Priming Task; IST = Internal Shift Task; CCT = Cognitive Control Training; PVT = Peripheral Vision Training; BATD = Brief Behavioral Activation Therapy for Depression

 

 

68

Table 2. Participant characteristics of the Intent-to-Treat sample. All participants (N = 34)

CCT Group (n = 21)

PVT Group (n = 13)

52 (18)

47.6 (10)

61.5 (8)

35.6 (14.6)

36.3 (14.4)

34.38 (15.4)

Caucasian, % (n)

73.5 (25)

85.7 (18)

53.8 (7)

African American, % (n)

20.6 (7)

9.5 (2)

38.5 (5)

Asian, % (n)

0 (0)

0 (0)

0 (0)

American Indian or Alaskan Native % (n)

0 (0)

0 (0)

0 (0)

Native Hawaiian or Pacific Islander % (n)

0 (0)

0 (0)

0 (0)

5.9 (2)

4.8 (1)

7.7 (1)

5.9 (2)

4.8 (1)

7.7 (1)

94.1 (32)

95.2 (20)

92.3 (12)

No degree

2.9 (1)

4.8 (1)

0 (0)

High School degree

11.8 (4)

9.5 (2)

15.4 (2)

Some college

29.4 (10)

19.0 (4)

46.2 (6)

Bachelor’s degree

26.5 (9)

28.6 (6)

23.1 (3)

Graduate training

29.4 (10)

38.1 (8)

15.4 (2)

Taking psychiatric medication, % (n)

29.4 (10)

28.6 (6)

30.8 (4)

BDI, Mean (SD)

29.6 (10.1)

28.8 (9.6)

30.8 (11.0)

MADRS, Mean (SD)

26.6 (7.6)

26.9 (8.2)

26.2 (6.9)

RRS, Mean (SD)

59.5 (9.9)

59.1 (7.6)

57.6 (13.2)

RPI, Mean (SD)

43.3 (7.4)

43.1 (6.9)

43.8 (8.4)

Sex, % female (n) Age, Mean (SD) Race

Other % (n) Ethnicity Hispanic % (n) Non-Hispanic % (n) Education

Note. BDI = Beck Depression Inventory-II, MADRS = Montgomery-Asberg Depression Rating Scale; RRS = Ruminative Response Scale; RPI = Reward Probability Index

 

 

69

Table 3. Participant characteristics of the completer sample. All Study Completers (n = 26)

CCT Group Completers (n = 14)

PVT Group Completers (n = 12)

61.5 (16)

64.3 (9)

58.3 (7)

35.5 (14.7)

37.2 (14.0)

33.6 (15.8)

Caucasian, % (n)

73.1 (19)

85.7 (12)

58.3 (7)

African American, % (n)

19.2 (5)

7.1 (1)

33.3 (4)

Asian, % (n)

0 (0)

0 (0)

0 (0)

American Indian or Alaskan Native % (n)

0 (0)

0 (0)

0 (0)

Native Hawaiian or Pacific Islander % (n)

0 (0)

0 (0)

0 (0)

7.7 (2)

7.1 (1)

8.3 (1)

7.7 (2)

7.1 (1)

8.3 (1)

92.3 (24)

92.9 (13)

91.7 (11)

0 (0)

0 (0)

0 (0)

High School degree

11.5 (3)

7.1 (1)

16.7 (2)

Some college

30.8 (8)

21.4 (3)

41.7 (5)

Bachelor’s degree

30.8 (8)

35.7 (5)

25.0 (3)

Graduate training

26.9 (7)

35.7 (5)

16.7 (2)

Taking psychiatric medication, % (n)

34.6 (9)

35.7 (5)

33.3 (4)

BDI, Mean (SD)

29.4 (9.1)

27.8 (6.6)

31.3 (11.3)

MADRS, Mean (SD)

26.2 (6.4)

25.6 (6.3)

27.0 (6.6)

RRS, Mean (SD)

58.9 (10.7)

59.3 (8.2)

58.5 (13.4)

Sex, % female (n) Age, Mean (SD) Race

Other % (n) Ethnicity Hispanic % (n) Non-Hispanic % (n) Education No degree

RPI, Mean (SD)

44.1 (6.9) 44.7 (5.3) 43.3 (8.6) Note. BDI = Beck Depression Inventory-II, MADRS = Montgomery-Asberg Depression Rating Scale; RRS = Ruminative Response Scale; RPI = Reward Probability Index

 

 

70

Table 4. Participant characteristics by completer and dropout status.

Sex, % female (n) Age, Mean (SD)

Completers (n = 26)

Dropouts (n = 8)

61.5 (16)

25.0 (2)

35.5 (14.7)

35.6 (15.2)

73.1 (19)

75.0 (6)

19.3 (5)

25.0 (2)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

0 (0)

7.6 (2)

0 (0)

92.3 (24)

100.0 (8)

Race Caucasian, % (n) African American, % (n) Asian, % (n) American Indian or Alaskan Native % (n) Native Hawaiian or Pacific Islander % (n) Other % (n) Ethnicity Hispanic % (n) Non-Hispanic % (n) Education No degree

0

12.5 (1)

High School degree

11.5 (3)

12.5 (1)

Some college

30.8 (8)

25 (2)

Bachelor’s degree

30.8 (8)

12.5 (1)

Graduate training

26.9 (7)

37.5 (3)

Taking psychiatric medication, % (n)

34.6 (9)

12.5 (1)

BDI, Mean (SD)

29.4 (9.1)

30.0 (13.5)

MADRS, Mean (SD)

26.2 (6.4)

27.9 (11.2)

RRS, Mean (SD)

58.9 (10.7)

57.1 (7.4)

RPI, Mean (SD)

44.1 (6.9)

40.9 (8.8)

CCT, % (n)

53.8 (14)

87.5 (7)

PVT, % (n)

46.2 (12)

12.5 (1)

Treatment Condition

Note. BDI = Beck Depression Inventory-II, MADRS = Montgomery-Asberg Depression Rating Scale; RRS = Ruminative Response Scale; RPI = Reward Probability Index; CCT = Cognitive Control Training; PVT = Peripheral Vision Training

 

 

71

Table 5. Primary and secondary outcome measures for all randomized participants at baseline, mid-treatment, post-treatment, and follow-up (N = 34). Baseline

Mid-treatment

Post-treatment

Follow-up

PVT

30.7 (11.0)

24.0 (12.5)

18.6 (13.3)

18.2 (14.6)

CCT

28.8 (9.6)

23.9 (11.3)

23.1 (10.8)

22.0 (11.8)

PVT

26.2 (6.9)

--

17.00 (9.9)

17.5 (11.7)

CCT

26.9 (8.2)

--

19.76 (11.1)

20.0 (9.8)

PVT

57.6 (13.2)

56.2 (16.2)

48.9 (15.2)

44.7 (14.3)

CCT

59.1 (7.6)

57.2 (8.7)

54.6 (12.2)

52.0 (11.2)

PVT

13.2 (3.4)

13.1 (4.2)

11.7 (3.7)

10.2 (3.9)

CCT

12.9 (2.4)

12.1 (2.7)

12.1 (3.7)

11.7 (3.2)

PVT

23.7 (5.4)

24.2 (6.6)

27.0 (6.4)

27.9 (7.2)

CCT

22.9 (5.0)

22.9 (5.8)

26.2 (6.7)

25.1 (6.7)

PVT

20.1 (4.6)

21.2 (5.8)

23.2 (5.5)

23.5 (5.4)

CCT

20.2 (4.4)

19.2 (4.5)

20.5 (5.7)

21.0 (5.1)

PVT

17.8 (9.4)

--

11.6 (7.4)

13.2 (11.4)

CCT

15.8 (8.5)

--

13.7 (8.9)

12.8 (9.3)

BDI

MADRS

RRS total

RRS-Brooding

RPI-RP

RPI-ES

BAI

Note. CCT = Cognitive Control Training; PVT = Peripheral Vision Training; BDI = Beck Depression Inventory-II, MADRS = Montgomery-Asberg Depression Rating Scale; RRS total = Ruminative Response Scale total; RRS-Brooding = Brooding subscale of Ruminative Response Scale; RPI-RP = Reward Probability Index – Reward Probability subscale; RPI-ES = Reward Probability Index – Environmental Suppressors subscale; BAI = Beck Anxiety Inventory

 

 

72

Table 6. Primary and secondary outcome measures for all treatment completers at baseline, mid-treatment, post-treatment, and follow-up (n = 26). Baseline

Mid-treatment

Post-treatment

Follow-up

PVT

31.3 (11.3)

23.9 (13.0)

18.1 (13.7)

17.6 (15.1)

CCT

27.8 (6.9)

22.1 (9.0)

20.9 (7.5)

19.2 (9.1)

PVT

27.0 (6.6)

--

17.0 (10.4)

17.6 (12.3)

CCT

25.5 (6.6)

--

17.8 (9.0)

18.2 (6.3)

PVT

58.5 (13.4)

56.9 (16.7)

49.0 (15.9)

44.5 (14.9)

CCT

59.8 (8.6)

57.3 (10.3)

53.3 (14.1)

49.1 (11.3)

PVT

13.5 (3.3)

13.4 (4.2)

11.9 (3.9)

10.3 (4.1)

CCT

13.1 (2.5)

11.9 (2.9)

12.1 (4.1)

11.4 (3.2)

PVT

23.4 (5.5)

23.9 (6.8)

27.0 (6.7)

27.9 (8.1)

CCT

23.8 (3.6)

23.6 (5.4)

28.4 (6.6)

26.3 (7.0)

PVT

19.9 (4.8)

21.1 (6.0)

23.3 (5.7)

23.7 (5.6)

CCT

20.9 (3.3)

19.1 (3.7)

20.8 (5.7)

21.6 (4.4)

PVT

18.3 (9.6)

--

11.6 (7.8)

13.3 (11.9)

CCT

15.6 (8.7)

--

12.7 (8.5)

11.2 (9.0)

BDI

MADRS

RRS total

RRS-Brooding

RPI-RP

RPI-ES

BAI

Note. CCT = Cognitive Control Training; PVT = Peripheral Vision Training; BDI = Beck Depression Inventory-II, MADRS = Montgomery-Asberg Depression Rating Scale; RRS total = Ruminative Response Scale total; RRS-Brooding = Brooding subscale of Ruminative Response Scale; RPI-RP = Reward Probability Index – Reward Probability subscale; RPI-ES = Reward Probability Index – Environmental Suppressors subscale; BAI = Beck Anxiety Inventory

 

 

73

Table 7. Response and remission rates by treatment condition. Intent-to-treat Sample

Completer Sample

CCT

PVT

ChiSquare

CCT

PVT

ChiSquare

% responders (n)

14.3 (3)

38.5 (5)

2.61

21.4 (3)

41.7 (5)

1.24

% remitters (n)

9.5 (2)

30.8 (4)

2.49

14.3 (2)

33.3 (4)

1.32

% responders (n)

23.8 (5)

38.5 (5)

0.83

30.8 (4)

41.7 (5)

0.32

% remitters (n)

19.0 (4)

38.5 (5)

1.56

30.8 (4)

41.7 (5)

0.32

Week 5

Week 8

Note. CCT = Cognitive Control Training; PVT = Peripheral Vision Training

 

 

74

Table 8. Means and Standard Deviations (SDs) for NAP task response times and NAP scores over time. Week 1

Week 3 (midtreatment)

Week 5 (posttreatment)

Week 8 (follow-up)

Control – negative RT

874.21 (110.88)

841.50 (109.55)

824.41 (137.58)

799.27 (106.34)

Control positive RT

878.48 (105.18)

816.82 (90.87)

798.91 (90.98)

796.26 (118.36)

NP – negative RT

911.26 (144.72)

868.81 (108.70)

837.28 (143.55)

818.30 (101.80)

NP – positive RT

899.10 (120.92)

856.42 (112.13)

813.99 (102.81)

828.34 (106.82)

NAP score negative

37.05 (98.88)

27.31 (72.82)

12.87 (77.74)

19.03 (65.24)

NAP score positive

20.61 (74.22)

39.60 (69.10)

15.08 (63.01)

32.08 (55.90)

Note. Control negative RT = Response time (in ms) for control trials with negative stimuli as target Control – positive RT =Response time (in ms) for control trials with positive stimuli as target NP- negative RT = Response time (in ms) for negative priming trials with negative stimuli as target NP - positive RT = Response time (in ms) for negative priming trials with positive stimuli as target NAP score - negative = difference between mean RT for Control-negative and Negative primingnegative trials (higher score represents greater inhibitory control of negative stimuli) NAP score - positive: difference between mean RT for Control-positive and Negative primingpositive trials (higher score represents greater inhibitory control of positive stimuli)

 

 

75

Table 9. Hierarchical multiple regression analyses of the predictive influence of baseline NAP-negative scores and treatment condition on treatment outcomes at follow-up. B BDI week 8 Step 1 BDI week 1 Step 2 NAP-negative score week 1 Treatment Condition Step 3 NAP-negative X Condition MADRS week 8 Step 1 MADRS week 1 Step 2 NAP-negative score week 1 Treatment Condition Step 3 NAP-negative X Condition RRS-brooding week 8 Step 1 RRS total week 1 Step 2 NAP-negative score week 1 Treatment Condition Step 3 NAP-negative X Condition

t

p

Model R2 .32

.73

3.11*

.01 .33

-.01 1.49

-.06 .64

.76 .53

-.03

-1.18

.26

.38

.43 .96

4.01**

.00 .45

-.01 1.15

-.56 .69

.58 .50 .54

-.03

-1.86

.08

.33 .78

3.23*

.00

.01 .26

2.28* .44

.04 .67

.49

.55 -.01

-1.64

.12

RRS-reflection week 8 Step 1 .52 RRS total week 1 .64 4.74** .00 Step 2 .55 NAP-negative score week 1 .01 1.11 .28 Treatment Condition .22 .44 .67 Step 3 .57 NAP-negative X Condition -.01 -.86 .40 Note. BDI = Beck Depression Inventory-II; NAP = Negative Affective Priming task; MADRS = Montgomery-Asberg Depression Rating Scale; RRS-brooding = Brooding subscale of the

 

  Ruminative Response Scale; RRS-reflection = Reflection subscale of the Ruminative Response Scale * p < .05. ** p < .001.

 

76

 

77

Table 10. Means and standard deviations for response time on Internal Shift Task (IST) Trials Baseline Mean (SD)

Week 3 Mean (SD)

Week 5 Mean (SD)

Week 8 Mean (SD)

Emotion Condition Non-Switch

1249.29 (421.51)

1006.78 (292.70)

956.59 (306.42)

956.13 (207.30)

Switch

1584.35 (488.32)

1353.69 (416.75)

1224.27 (353.36)

1230.25 (291.98)

Gender Condition Non-Switch

1052.17 (314.09)

921.58 (229.04)

872.75 (191.72)

844.10 (193.96)

Switch

1419.70 (406.10)

1234.66 (342.44)

1108.42 (300.25)

1082.23 (325.58)

335.06 (191.80)

346.92 (187.06)

267.68 (193.45)

274.13 (186.28)

367.52 (217.68)

313.08 (230.25)

235.67 (193.80)

238.125 (218.28)

Emotion Switch Cost Gender Switch Cost

Note. All results reported in milliseconds. Emotion and Gender Switch Cost scores were calculated as the difference between average response time for switch and non-switch trials in the emotion and gender conditions, respectively.

 

 

78

Table 11. Hierarchical multiple regression analyses of the predictive influence of baseline IST emotion switch cost and treatment condition on week 8 treatment outcomes. BDI week 8 Step 1 BDI week 1 Step 2 Emotion switch cost week 1 Treatment Condition Step 3 Condition X Emotion Switch cost MADRS week 8 Step 1 MADRS week 1 Step 2 Emotion switch cost week 1 Treatment Condition Step 3 Condition X Emotion Switch cost RRS-brooding week 8 Step 1 RRS-brooding week 1 Step 2 Emotion switch cost week 1 Treatment Condition Step 3 Condition X Emotion Switch cost

B

t

p

.69

2.71*

.01

Model R2 .26 .29

-.01 2.12

-0.55 0.87

.59 .40

.02

1.13

.27

.34

.43 1.00

4.00**

.00 .44

.00 .58

0.27 0.35

.79 .73 .45

.01

0.67

.51

.24 .60

2.61*

.02

.00 .93

0.28 1.41

.78 .18

.32

.39 .01

1.48

.16

RRS-reflection week 8 Step 1 .43 RRS-reflection week 1 .66 3.99** .00 Step 2 .54 Emotion switch cost week 1 -.01 -1.75 .10 Treatment Condition .89 1.65 .12 Step 3 .55 Condition X Emotion Switch cost .00 0.69 .50 Note. BDI = Beck Depression Inventory-II; NAP = Negative Affective Priming task; MADRS = Montgomery-Asberg Depression Rating Scale; RRS-brooding = Brooding subscale of

 

  the Ruminative Response Scale; RRS-reflection = Reflection subscale of the Ruminative Response Scale * p < .05. ** p < .001

 

79

 

80 FIGURES Figure 1. CONSORT diagram of participant enrollment.

Assessed for eligibility by phone (n = 78)  

Excluded (n = 26) Reasons for exclusion: Not depressed (n=7) Bipolar Disorder or Psychosis (n = 11) Use of exclusionary medication (n=5) Current alcohol or drug dependence (n = 2) Over age 65 (n = 1)

Scheduled for screening visit (n = 52)

Did not attend screening visit (n = 9)

Excluded (n = 6) Reasons for Exclusion:

Eligible (n = 37)

Did not meet criteria for MDD (n = 3) Diagnosis of bipolar disorder (n = 2) Principal diagnosis other than MDD (n = 1)

Lost to follow up prior to treatment assignment (n = 3)

 

Allocated to BA + CCT (n = 21)

Allocated to BA + Control (PVT) (n = 13)

Completed treatment (n = 14)

Completed treatment (n = 12)

Dropped out (n = 7)

Dropped out (n =1)

 

81

Figure 2. Negative Affective Priming Task: Example of a priming trial (in the negative information condition)

   

500 ms

+    

Prime Trial Defeated Admired    

500 ms

+    

Test Trial Elated Troubled

Note. In a negative priming trial, the prime trial distractor (the red word) shares the same valence as the test trial target (the blue word). The negative priming trial can be negative (i.e., the prime trial distractor and test trial targets are negatively valenced) or positive (i.e., the prime trial distractor and test trial targets are positively valenced).

 

 

82

Figure 3. Negative Affective Priming Task: Example of a control trial.

   

500 ms

+    

Prime Trial Quiet Shamed    

500 ms

+    

Test Trial Hurt Admired

Note. In a control condition trial, the prime trial distractor (the red word) is neutrally valenced and does not share the same valence as the test trial target (the blue word, which is either negative or positive). The control trial can be negative (i.e., the test trial target is negatively valenced) or positive (i.e., the test trial targets is positively valenced).

 

 

83

Figure 4. NAP-Positive Scores over time by treatment condition.

50

NAP positive scores (ms)

40 30 20 10 0 -10 -20

Baseline

Week 5

Week 8

CCT

37.5911

-12.29

24.9467

PVT

-1.7691

37.4764

33.9127

Note. CCT = Cognitive Control Training; PVT = Peripheral Vision Training

 

 

84 REFERENCES

Abela, J.Z., & Hankin, B.L. (2011). Rumination as a vulnerability factor to depression during the transition from early to middle adolescence: A multiwave longitudinal study. Journal of Abnormal Psychology, 120, 259-271. doi:10.1037/a0022796 Addis, M. E., & Carpenter, K. M. (1999). Why, why, why?: Reason-giving and rumination as predictors of response to activation-and insight-oriented treatment rationales. Journal of Clinical Psychology, 55, 881-894. doi: 10.1002/(SICI)10974679(199907)55:73.0.CO;2-E American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC. APA Presidential Task Force on Evidence-Based Practice. (2006). Evidence-based practice in psychology. American Psychologist, 61, 271-285. doi: 10.1037/0003066X.61.4.271 Baron, R.M., & Kenny, D.A. (1986). The moderator-mediator variable distinction in social psychological research: conceptual, strategic, and statistical considerations. Journal of Personality & Social Psychology, 51, 1173-1182. doi: 10.1037/00223514.51.6.1173 Barrera M. Jr. (1979). An evaluation of a brief group therapy for depression. Journal of Consulting and Clinical Psychology, 47, 413-415. Beck, A.T., & Steer, R.A. (1993). Beck Anxiety Inventory: Manual (2nd ed.). San Antonio, TX: The Psychological Corporation. Beck, A.T., Steer, R.A., & Brown, G. (1996). Beck Depression Inventory Second Edition Manual. San Antonio: The Psychological Corporation. Beck, A.T., Steer, R.A., & Garbin, M.G. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of research. Clinical Psychology Review, 8, 77-100. doi:10.1016/0272-7358(88)90050-5 Beckwé, M., Deroost, N., Koster, E. H., De Lissnyder, E., & De Raedt, R. (2013). Worrying and rumination are both associated with reduced cognitive control. Psychological research, 1-10. doi: 10.1007/s00426-013-0517-5 Benedetti, F., Mayberg, H. S., Wager, T. D., Stohler, C. S., & Zubieta, J. K. (2005). Neurobiological mechanisms of the placebo effect. The Journal of Neuroscience, 25(45), 10390-10402. doi: 10.1523/JNEUROSCI.3458-05.2005  

 

85

Bernblum, R., & Mor, N. (2010). Rumination and emotion-related biases in refreshing information. Emotion, 10, 423-432. doi:10.1037/a0018427 Brown, T.A., DiNardo, P.A., & Barlow, D.H. (1994). Anxiety Disorders Interview Schedule for DSM-IV. New York, NY: Graywind Publications, Inc. Brunoni, A. R., Boggio, P. S., De Raedt, R., Benseñor, I. M., Lotufo, P. A., Namur, V., ... & Vanderhasselt, M. A. (2014). Cognitive control therapy and transcranial direct current stimulation for depression: A randomized, double-blinded, controlled trial. Journal of Affective Disorders, 162, 43-49. doi:10.1016/j.jad.2014.03.026 Burns, D.D., Sayers, S.L., & Moras, K. (1994). Intimate relationships and depression: is there a causal connection? Journal of Consulting & Clinical Psychology, 62, 1033-1043. doi:10.1037/0022-006X.62.5.1033 Busch, A. M., Uebelacker, L. A., Kalibatseva, Z., & Miller, I. W. (2010). Measuring homework completion in behavioral activation. Behavior Modification, 34, 310329. doi: 10.1177/0145445510373384 Butler, L. D., & Nolen-Hoeksema, S. (1994). Gender differences in responses to depressed mood in a college sample. Sex Roles, 30, 331-346. doi:                         10.1007/BF01420597 Calkins A.W., McMorran K.E., Siegle, G.J., & Otto, M.W. (2014). The effects of computerized cognitive control training on community adults with depressed mood. Behavioural and Cognitive Psychotherapy, 3, 1-12. doi:10.1017/S1352465814000046 Carlbring, P., Hägglund, M., Luthström, A., Dahlin, M., Kadowaki, Å., Vernmark, K., & Andersson, G. (2013). Internet-based behavioral activation and acceptance-based treatment for depression: a randomized controlled trial. Journal of Affective Disorders, 148, 331-337. doi:10.1016/j.jad.2012.12.020 Carvalho, J. P., Gawrysiak, M. J., Hellmuth, J. C., McNulty, J. K., Magidson, J. F., Lejuez, C. W., & Hopko, D. R. (2011). The reward probability index: design and validation of a scale measuring access to environmental reward. Behavior Therapy, 42, 249-262. doi: 10.1016/j.beth.2010.05.004. Chen, Y.W., & Dilsaver, S.C. (1996). Lifetime rates of suicide attempts among subjects with bipolar and unipolar disorders relative to subjects with other Axis I disorders. Biological Psychiatry, 39, 896-899. doi: 0006-3223(95)00295-2 Chermack, S. T., Zivin, K., Valenstein, M., Ilgen, M., Austin, K. L., Wryobeck, J., &  

 

86 Blow, F. C. (2008). The prevalence and predictors of mental health treatment services in a national sample of depressed veterans. Medical Care, 46, 813-820. doi: 10.1097/MLR.0b013e318178eb08

Ciesla, J.A., & Roberts, J.E. (2002). Self-directed thought and response to treatment for depression: a preliminary investigation. Journal of Cognitive Psychotherapy, 16, 435–453. doi:10.1891/jcop.16.4.435.52528 Cohen, J (1988) Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum. Cooper, A. E., Corrigan, P. W., & Watson, A. C. (2003). Mental illness stigma and care seeking. The Journal of Nervous and Mental Disease, 191, 339-341. Cornwall, P. L., & Scott, J. (1997). Partial remission in depressive disorders. Acta Psychiatrica Scandinavica, 95, 265-271. doi:  10.1111/j.16000447.1997.tb09630.x Corrigan, P., Markowitz, F. E., Watson, A., Rowan, D., & Kubiak, M. A. (2003). An attribution model of public discrimination towards persons with mental illness. Journal of Health and Social Behavior, 44, 162-179. Curran, P.J., Obeidat, K., & Losardo, D. (2010). Twelve frequently asked questions about latent growth curve modeling. Journal of Cognition and Development, 11, 121136. doi:  10.1080/15248371003699969 Daughters, S. B., Braun, A. R., Sargeant, M. N., Reynolds, E. K., Hopko, D. R., Blanco, C., & Lejuez, C. W. (2008). Effectiveness of a brief behavioral treatment for inner-city illicit drug users with elevated depressive symptoms: the life enhancement treatment for substance use (LETS Act!). Journal of Clinical Psychiatry, 69, 122-129. Davidson, J., Turnbull, C. D., Strickland, R., Miller, R. & Graves, K. (1986). The Montgomery-Åsberg Depression Scale: Reliability and validity. Acta Psychiatrica Scandinavica, 73, 544–548. doi: 10.1111/j.1600-0447.1986.tb02723.x Davidson, R. J. (1994). Asymmetric brain function affective style and psychopathology: The role of early experience and plasticity. Development and Psychopathology, 6, 741–758. doi:10.1017/S0954579400004764 Davis, R. N., & Nolen-Hoeksema, S. (2000). Cognitive inflexibility among ruminators and nonruminators. Cognitive Therapy and Research, 24, 699-711. doi: 10.1023/A:1005591412406  

 

87

De Lissnyder, E., Koster, E. H., & De Raedt, R. (2012). Emotional interference in working memory is related to rumination. Cognitive Therapy and Research, 36, 348-357. doi:10.1007/s10608-011-9352-4. De Lissnyder, E., Koster, E. H., Everaert, J., Schacht, R., Van den Abeele, D., & De Raedt, R. (2012). Internal cognitive control in clinical depression: General but no emotion-specific impairments. Psychiatry Research, 199, 124-130. doi:   10.1016/j.psychres.2012.04.019. Demeyer, I., De Lissnyder, E., Koster, E. H., & De Raedt, R. (2012). Rumination mediates the relationship between impaired cognitive control for emotional information and depressive symptoms: A prospective study in remitted depressed adults. Behavior Research and Therapy, 50, 292-297. doi: S0005-7967(12)000435 Dimidjian, S., Hollon, S. D., Dobson, K. S., Schmaling, K. B., Kohlenberg, R. J., Addis, M. E., ... & Jacobson, N. S. (2006). Randomized trial of behavioral activation, cognitive therapy, and antidepressant medication in the acute treatment of adults with major depression. Journal of Consulting and Clinical Psychology, 74, 658. doi:10.1037/0022-006X.74.4.658 Dimidjian, S., Barrera Jr, M., Martell, C., Muñoz, R. F., & Lewinsohn, P. M. (2011). The origins and current status of behavioral activation treatments for depression. Annual Review of Clinical Psychology, 7, 1-38. doi:10.1146/annurevclinpsy-032210-104535 Donaldson, C., Lam, D., & Mathews, A. (2007). Rumination and attention in major depression. Behavior Research & Therapy, 45, 2664–2678. doi:10.1016/j.brat.2007.07.002 Drevets, W. C. (1999). Prefrontal cortical amygdalar metabolism in major depression. Annals of the New York Academy of Sciences, 877, 614–637. doi:  10.1111/j.17496632.1999.tb09292.x Drevets W. C., Videen T., Price J., Preskorn S., Carmichael S., & Raichle M. A. (1992). A functional anatomical study of unipolar depression. Journal of Neuroscience, 12, 3628-3641. Duncan, T. E., & Duncan, S. C. (2004). An introduction to latent growth curve modeling. Behavior Therapy, 35, 333-363. doi:10.1016/S0005-7894(04)80042-X Ekers, D., Richards, D., McMillan, D., Bland, J.M., & Gilbody, S. (2011). Behavioural activation delivered by the non-specialist: phase II randomised controlled trial. The British Journal of Psychiatry, 198, 66-72. doi:10.1192/bjp.bp.110.079111  

 

88

Fales, C. L., Barch, D. M., Rundle, M. M., Mintun, M. A., Mathews, J., Snyder, A. Z., & Sheline, Y. I. (2009). Antidepressant treatment normalizes hypoactivity in dorsolateral prefrontal cortex during emotional interference processing in major depression. Journal of Affective Disorders, 112, 206-211. doi:10.1016/j.jad.2008.04.027 Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behavior Research Methods, 39, 175-191. doi:  10.3758/BF03193146 Fava, G., Ruini, C., & Belaise, C. (2006). The concept of recovery in major depression. Psychological Medicine, 37, 307. doi: 10.1017/s0033291706008981 Ferster, C. (1973). A functional analysis of depression. American Psychologist, 28, 857870. doi:10.1037/h0035605 First, M.B., Spitzer, R.L., Gibbon M., & Williams, J.B.W. (2002). Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Patient Edition. (SCID-I/P) New York: Biometrics Research, New York State Psychiatric Institute. Gawrysiak, M., Nicholas, C., & Hopko, D. R. (2009). Behavioral activation for moderately depressed university students: Randomized controlled trial. Journal of Counseling Psychology, 56, 468. http://dx.doi.org/10.1037/a0016383 Goeleven, E., De Raedt, R., Baert, S., & Koster, E. H. W. (2006). Deficient inhibition of emotional information in depression. Journal of Affective Disorders, 93, 149-157. doi:10.1016/j.jad.2006.03.007 Goeleven, E., De Raedt, R., Leyman, L., & Verschuere, B. (2008). The Karolinska directed emotional faces: a validation study. Cognition and Emotion, 22, 10941118. doi:10.1080/02699930701626582 Gotlib, I.H., & Joormann, J. (2010). Cognition and Depression: Current Status and Future Directions. Annual Review of Clinical Psychology, 6, 285-312. doi:10.1146/annurev.clinpsy.121208.131305 Gotlib, I. H., & Whiffen, V. E. (1989). Depression and marital functioning: an examination of specificity and gender differences. Journal of Abnormal Psychology, 98, 23-30. doi:10.1037/0021-843X.98.1.23 Gronwall, D. M. (1977). Paced auditory serial-addition task: A measure of recovery from  

 

 

89 concussion. Perceptual and Motor Skills, 44, 367-373. doi: 10.2466/pms.1977.44.2.367

Gros, D. F., & Haren, W. B. (2011). Open trial of brief behavioral activation psychotherapy for depression in an integrated veterans affairs primary care setting. The primary care companion to CNS disorders, 13(4). 10.4088/PCC.11m01136. Harvey P.O., Le Bastard G., Pochon J.B., Levy R., Allilaire J.F., Dubois, B., & Fossati, P. (2004). Executive functions and updating of the contents of working memory in unipolar depressions. Journal of Psychiatry Research, 38, 567–576. doi:10.1016/j.jpsychires.2004.03.003 Holtzheimer, P. E., & Mayberg, H. S. (2011). Stuck in a rut: rethinking depression and its treatment. Trends in Neurosciences, 34, 1-9. doi:10.1016/j.tins.2010.10.004 Hopko, D., Lejuez, C., LePage, J., Hopko, S., & McNeil, D. (2003). A brief behavioral activation treatment for depression: A randomized pilot trial within an inpatient psychiatric hospital. Behavior Modification, 27, 458-469. doi:10.1177/0145445503255489 Hopko, D., Bell, J., Armento, M., Hunt, M., & Lejuez, C. (2005). Behavior therapy for depressed cancer patients in primary care. Psychotherapy: Theory, Research, Practice, Training, 42, 236-243. doi:10.1037/0033-3204.42.2.236 Hopko, D. R., Armento, M. A., Robertson, S. C., Ryba, M. M., Carvalho, J. P., Colman, L. K., & ... Lejuez, C. W. (2011). Brief behavioral activation and problem-solving therapy for depressed breast cancer patients: Randomized trial. Journal of Consulting and Clinical Psychology, 79, 834-849. doi:10.1037/a0025450 Huttenlocher, J., Haight, W., Bryk, A., Seltzer, M., & Lyons, T. (1991). Early vocabulary growth: Relation to language input and gender. Developmental Psychology, 27, 236–248. doi:  10.1037/0012-1649.27.2.236 Jacobson, N. S., Dobson, K. S., Truax, P. A., Addis, M. E., Koerner, K., Gollan, J. K., Gortner, E., & Prince, S.E. (1996). A component analysis of cognitive– behavioral treatment for depression. Journal of Consulting and Clinical Psychology, 64, 295–304. doi:10.1037/0022-006X.64.2.295 Joormann, J. (2004). Attentional bias in dysphoria: The role of inhibitory processes. Cognition & Emotion, 18, 125-147. doi:10.1080/02699930244000480 Joormann, J. (2006). The relation of rumination and inhibition: evidence from a negative  

 

90 priming task. Cognitive Therapy and Research. 30, 149–60. doi:  10.1007/s10608006-9035-8

Joormann, J., Dkane, M., & Gotlib, I. H. (2006). Adaptive and maladaptive components of rumination? Diagnostic specificity and relation to depressive biases. Behavior Therapy, 37, 269-280. doi:10.1016/j.beth.2006.01.002 Joormann, J., & Gotlib, I. H. (2008). Updating the contents of working memory in depression: interference from irrelevant negative material. Journal of Abnormal Psychology, 117, 182-192. doi:10.1037/0021-843X.117.1.182 Joormann, J., & Gotlib, I. H. (2010). Emotion regulation in depression: Relation to cognitive inhibition. Cognition & Emotion, 24, 281-298. doi:10.1080/02699930903407948 Joormann, J., Levens, S. M., & Gotlib, I. H. (2011). Sticky thoughts: Depression and rumination are associated with difficulties manipulating emotional material in working memory. Psychological Science, 22, 979-983. doi:10.1177/0956797611415539 Joormann J., & Tran, T. (2009). Rumination and intentional forgetting of emotional material. Cognition & Emotion, 23, 1233–1246. doi:10.1080/02699930802416735 Khan, A., Khan, S. R., Shankles, E. B., & Polissar, N. L. (2002). Relative sensitivity of the Montgomery-Asberg Depression Rating Scale, the Hamilton Depression rating scale and the Clinical Global Impressions rating scale in antidepressant clinical trials. International clinical psychopharmacology, 17(6), 281-285. doi: 10.1097/00004850-200211000-00003 Kirsch, I., & Sapirstein, G. (1998). Listening to Prozac but hearing placebo: A metaanalysis of antidepressant medication. Prevention & Treatment, 1 (2), 2a. http://dx.doi.org/10.1037/1522-3736.1.1.12a Lazeron, R. H., Rombouts, S. A., de Sonneville, L., Barkhof, F., & Scheltens, P. (2003). A paced visual serial addition test for fMRI. Journal of Neurological Sciences, 213, 29–34. doi: 10.1016/S0022-510X(03)00144-8 Lewinsohn, P. (1974). A behavioral approach to depression. The psychology of depression: Contemporary theory and research. Oxford, UK: John Wiley. Lewinsohn, P.M., Munoz, R.F., Youngren, M.A., & Zeiss, A.M. (1978). Control Your Depression. Englewood Cliffs, NJ: Prentice-Hall.  

 

91

Lewinsohn, P.M., Sullivan, M.J., & Grosscup, S.J. (1980). Changing reinforcing events: an approach to the treatment of depression. Psychotherapy Theory Research and Practice, 17, 322–334. doi:  10.1037/h0085929 Lejuez, C. W., Hopko, D. R., Acierno, R., Daughters, S. B., & Pagoto, S. L. (2011). Ten year revision of the brief behavioral activation treatment for depression: revised treatment manual. Behavior Modification, 35, 111-161. doi:10.1177/0145445510390929 Lundqvist, D., Flykt, A., & Ohman, A. (1998). The averaged Karolinska directed emotional faces. Stockholm: Karolinska Institute, Department of Clinical Neuroscience, Section Psychology. Lyubomirsky, S., & Nolen-Hoeksema, S. (1995). Effects of self-focused rumination on negative thinking and interpersonal problem-solving. Journal of Personality and Social Psychology, 69, 176-190. doi:  http://dx.doi.org/10.1037/00223514.69.1.176 MacPherson, L., Tull, M. T., Matusiewicz, A. K., Rodman, S., Strong, D. R., Kahler, C. W., ... & Lejuez, C. W. (2010). Randomized controlled trial of behavioral activation smoking cessation treatment for smokers with elevated depressive symptoms. Journal of consulting and clinical psychology, 78, 55-61. doi:10.1037/a0017939. Magidson, J. F., Gorka, S. M., MacPherson, L., Hopko, D. R., Blanco, C., Lejuez, C. W., & Daughters, S. B. (2011). Examining the effect of the Life Enhancement Treatment for Substance Use (LETS ACT) on residential substance abuse treatment retention. Addictive Behaviors, 36, 615-623. doi:10.1016/j.addbeh.2011.01.016 Mayberg, H.S. (1997). Limbic-cortical dysregulation: a proposed model of depression. Journal of Neuropsychiatry & Clinical Neuroscience, 9, 471– 481. Mayberg, H.S., Brannan, S.K., Tekell, J.L., Silva, A., Mahurin, R.K., McGinnis, S., & Jerabek, P.A. (2000). Regional metabolic effects of fluoxetine in major depression: Serial changes and relation- ship to clinical response. Biological Psychiatry, 48, 830–843. doi:10.1016/S0006-3223(00)01036-2 Mayberg, H.S., Liotti, M., Brannan, S. K., McGinnis, S., Mahurin, R. K., Jerabek, P. A., . . . Fox, P.T. (1999). Reciprocal limbic-cortical function and negative mood: Converging PET findings in depression and normal sadness. American Journal of Psychiatry, 156, 675–682.  

 

92

Merriam, E. P., Thase, M. E., Haas, G. L., Keshavan, M. S., & Sweeney, J. A. (1999). Prefrontal cortical dysfunction in depression determined by Wisconsin Card Sorting Test performance. American Journal of Psychiatry, 156, 780–782. Michalak, J., Hölz, A., & Teismann, T. (2011). Rumination as a predictor of relapse in mindfulness-based cognitive therapy for depression. Psychology and Psychotherapy: Theory, Research and Practice, 84, 230-236. doi:   10.1348/147608310X520166 Miranda, R., & Nolen-Hoeksema, S. (2007) Brooding and reflection: Rumination predicts suicidal ideation at 1-year follow-up in a community sample. Behavior Research and Therapy, 45, 3088-3095. doi:10.1016/j.brat.2007.07.015 Mojtabai, R. (2009). Unmet need for treatment of major depression in the United States. Psychiatric Services, 60, 297-305. 10.1176/appi.ps.60.3.297. Montgomery, S. A., & Asberg, M. (1979). A new depression scale designed to be sensitive to change. British Journal of Psychiatry, 134, 382-389. doi:10.1192/bjp.134.4.382 Moshier, S.J., Molokotos, E.K., Stein, A.T., & Otto, M.W. (in press). Assessing the effects of depressed mood and cognitive control training on memory confidence and accuracy following repeated checking. International Journal of Cognitive Therapy. Murphy, K.R., & Myors, B. (2004). Statistical power analysis: A Simple and General Model for Traditional and Modern Hypothesis Tests (2nd ed.). Lawrence Erlbaum, Mahwah, NJ. Muthen, B.O., & Curran, P. (1997). General longitudinal modeling of individual differences in experimental designs: A latent variable framework for analysis and power estimation. Psychological Methods, 2, 371-402. doi:   http://dx.doi.org/10.1037/1082-989X.2.4.371 Muthen, L.K., & Muthen, B.O. (2002). How to use a monte carlo study to decide on sample size and determine power. Structural Equation Modeling, 4, 599-620. doi:0.1207/S15328007SEM0904_8 Nolen-Hoeksema, S. (2000). The role of rumination in depressive disorders and mixed anxiety/depressive symptoms. Journal Of Abnormal Psychology, 109, 504-511. doi:10.1037/0021-843X.109.3.504 Nolen-Hoeksemam S., Wisco, B.E., & Lyubomirsky, S. (2008). Rethinking rumination.  

 

93 Perspectives in Psychological Science, 3, 400–424. doi:  10.1111/j.17456924.2008.00088.x

Nolen-Hoeksema, S., & Morrow, J. (1991). A prospective study of depression and posttraumatic stress symptoms after a natural disaster: the 1989 Loma Prieta earthquake. Journal of Personality and Social Psychology, 61, 115-121. doi:   http://dx.doi.org/10.1037/0022-3514.61.1.115 Nolen-Hoeksema, S., Morrow, J., & Fredrickson, B. L. (1993). Response styles and the duration of episodes of depressed mood. Journal of Abnormal Psychology, 102, 20-28. doi:10.1037/0021-843X.102.1.20 Ochsner, K. N., & Gross, J. J. (2008). Cognitive emotion regulation insights from social cognitive and affective neuroscience. Current Directions in Psychological Science, 17, 153-158. doi:  10.1111/j.1467-8721.2008.00566.x Ochsner, K. N., Ray, R. D., Cooper, J. C., Robertson, E. R., Chopra, S., & Gabrieli, J. D., & Gross, J.J. (2004). For better or for worse: Neural systems supporting the cognitive down- and up-regulation of negative emotion. Neuroimage, 23, 483– 499.  doi:10.1016/j.neuroimage.2004.06.030 Onraedt, T., & Koster, E. H. (2014). Training working memory to reduce rumination. PloS one, 9, e90632. doi:  0.1371/journal.pone.0090632 O'Mahen, H. A., Woodford, J., McGinley, J., Warren, F. C., Richards, D. A., Lynch, T. R., & Taylor, R. S. (2013). Internet-based behavioral activation—Treatment for postnatal depression (Netmums): A randomized controlled trial. Journal of Affective Disorders, 150, 814-822. doi: 10.1016/j.jad.2013.03.005 Otto, M. W., Teachman, B. A., Cohen, L. S., Soares, C. N., Vitonis, A. F., & Harlow, B. L. (2007). Dysfunctional attitudes and episodes of major depression: Predictive validity and temporal stability in never-depressed, depressed, and recovered women. Journal of Abnormal Psychology, 116, 475-483. http://dx.doi.org/10.1037/0021-843X.116.3.475 Phillips, M. L., Drevets, W. C., Rauch, S. L., & Lane, R. (2003). Neurobiology of emotion perception II: Implications for major psychiatric disorders. Biological Psychiatry, 54, 515–528.  doi:10.1016/S0006-3223(03)00171-9 Prins, M. A., Verhaak, P. F., Smolders, M., Laurant, M. G., van der Meer, K., Spreeuwenberg, P., ... & Bensing, J. M. (2010). Patient factors associated with guideline-concordant treatment of anxiety and depression in primary care. Journal of General Internal Medicine, 25, 648-655. doi:  10.1007/s11606-009-1216-1  

 

94

Richter, P., Werner, J., Heerlein, A., Kraus, A., & Sauer, H. (1998). On the validity of the Beck Depression Inventory. Psychopathology, 31, 160–168. doi: 10.1159/000066239 Ritchey, M., Dolcos, F., Eddington, K. M., Strauman, T. J., & Cabeza, R. (2011). Neural correlates of emotional processing in depression: changes with cognitive behavioral therapy and predictors of treatment response. Journal of Psychiatric Research, 45, 577-587. doi:10.1016/j.jpsychires.2010.09.007 Rovner, B. W., German, P. S., Brant, L. J., Clark, R., Burton, L., & Folstein, M. F. (1991). Depression and mortality in nursing homes. JAMA: Journal of the American Medical Association, 265, 993-996. Doi:10.1001/jama.1991.03470020041019. Schotte, C. K., Maes, M., Cluydts, R., De Doncker, D., & Cosyns, P. (1997). Construct validity of the Beck Depression Inventory in a depressive population. Journal of Affective Disorders, 46, 115-125. doi: S0165-0327(97)00094-3 Schmaling, K. B., Dimidjian, S., Katon, W., & Sullivan, M. (2002). Response styles among patients with minor depression and dysthymia in primary care. Journal of Abnormal Psychology, 111, 350-356. doi:10.1037/0021-843X.111.2.350 Schoofs, H., Hermans, D., & Raes, F. (2010). Brooding and reflection as subtypes of rumination: Evidence from confirmatory factor analysis in nonclinical samples using the Dutch Ruminative Response Scale. Journal of Psychopathology and Behavioral Assessment, 32, 609-617. doi:10.1007/s10862-010-9182-9 Segrave, R. A., Arnold, S., Hoy, K., & Fitzgerald, P. B. (2014). Concurrent cognitive control training augments the antidepressant efficacy of tDCS: a pilot study. Brain stimulation, 7, 325-331. doi:  10.1016/j.brs.2013.12.008. Sheline, Y.I., Barch, D.M., Donnelly, J.M., Ollinger, J.M., Snyder, A.Z., Mintun, M.A. (2001). Increased amygdala response to masked emotional faces in depressed subjects resolves with antidepressant treatment: an fMRI study. Biological Psychiatry, 50, 651-658. doi:10.1016/S0006-3223(01)01263-X Siegle, G. J. (1999). A neural network model of attention biases in depression. Progress in Brain Research, 121, 415–441. doi:10.1016/S0079-6123(08)63085-X Siegle, G. J., Steinhauer, S. R., Thase, M. E., Stenger, V. A., & Carter, C. S. (2002). Can’t shake that feeling: fMRI assessment of sustained amygdala activity in response to emotional information in depressed individuals. Biological Psychiatry, 51, 693–707. doi:10.1016/S0006-3223(02)01314-8  

 

95

Siegle, G.J., Thompson, W., Carter, C.S., Steinhauer, S.R., & Thase, M.E. (2007). Increased amygdala and decreased dorsolateral prefrontal BOLD responses in unipolar depression: related and independent features. Biological Psychiatry, 61, 198 - 209. doi:10.1016/j.biopsych.2006.05.048 Siegle, G.J., Ghinassi, F., & Thase, M.E. (2007). Neurobehavioral therapies in the 21st century: Summary of an emerging field and an extended example of Cognitive Control Training for depression. Cognitive Therapy and Research, 31, 235-262. doi:10.1007/s10608-006-9118-6 Siegle, G. J., Price, R. B., Jones, N. P., Ghinassi, F., Painter, T., & Thase, M. E. (2014). You gotta work at it: Pupillary indices of task focus are prognostic for response to a neurocognitive intervention for rumination in depression. Clinical Psychological Science, 2, 455-471. doi:  10.1177/2167702614536160 Smith, E. E., & Jonides, J. (1999). Storage and executive processes in the frontal lobes. Science, 283, 1657-1661. doi:10.1126/science.283.5408.1657 Solomon, D. A., Keller, M. B., Leon, A. C., Mueller, T. I., Lavori, P. W., Shea, M. T., . . . & Endicott, J. (2000). Multiple recurrences of major depressive disorder. American Journal of Psychiatry, 157, 229-233. doi:  10.1176/appi.ajp.157.2.229 Steer, R.A., Ranieri, W.F., Beck, A.T., & Clark, D.A. (1993). Further evidence for the validity of the Beck Anxiety Inventory with psychiatric outpatients. Journal of Anxiety Disorders, 7, 195-205. doi:10.1016/0887-6185(93)90002-3 Stewart, W. F., Ricci, J. A., Chee, E., Hahn, S. R., & Morganstein, D. (2003). Cost of lost productive work time among US workers with depression. JAMA: Journal of the American Medical Association, 289, 3135-3144. doi: 10.1001/jama.289.23.3135 Treynor, W., Gonzalez, R., & Nolen-Hoeksema, S.,(2003). Rumination reconsidered: A psychometric analysis. Cognitive Therapy & Research, 27, 247-259. doi:10.1023/A:1023910315561 Uher, R., Farmer, A., Maier, W., Rietschel, M., Hauser, J., Marusic, A., . . . Aitchison, K. J. (2008). Measuring depression: comparison and integration of three scales in the GENDEP study. Psychological Medicine, 38, 289-300. doi: S0033291707001730 Walsh, B. T., Seidman, S. N., Sysko, R., & Gould, M. (2002). Placebo response in studies of major depression: variable, substantial, and growing. JAMA, 287 (14), 1840-1847. doi:10.1001/jama.287.14.1840. Wells, A. (2000). Emotional disorders and metacognition: Innovative cognitive therapy.  

 

96 New York, NY US: John Wiley & Sons Ltd.

Winokur, G. (1997). All roads lead to depression: clinically homogeneous, etiologically heterogeneous. Journal of Affective Disorders, 45, 97-108. doi:10.1016/S01650327(97)00063-3 Witt, W. P., Keller, A., Gottlieb, C., Litzelman, K., Hampton, J., Maguire, J., & Hagen, E. W. (2011). Access to adequate outpatient depression care for mothers in the USA: a nationally representative population-based study. The Journal of Behavioral Health Services & Research, 38, 191-204. doi:10.1007/s11414-0099194-y. Whitmer, A., & Gotlib, I. H. (2011). Brooding and reflection reconsidered: A factor analytic examination of rumination in currently depressed, formerly depressed, and never depressed individuals. Cognitive Therapy and Research, 35, 99-107. doi: 10.1007/s10608-011-9361-3

 

 

97 CURRICULUM VITAE Samantha Moshier, M.A. Anxiety Disorders Center The Institute of Living 200 Retreat Avenue Hartford, CT 06114 Office phone: (860) 972–9806 Cell phone: (607) 215–3793 E-mail: [email protected]

EDUCATION 2014 – present

Predoctoral Internship in Clinical Psychology CBT Track The Institute of Living/Hartford Hospital Hartford, CT

September 2015 (Anticipated)

Ph.D. Candidate, Clinical Psychology Boston University Boston, Massachusetts Dissertation: “Cognitive Control Training as an Adjunct to Behavioral Activation Therapy in the Treatment of Depression” Advisor: Michael W. Otto, Ph.D.

May 2010

Master of Arts, Clinical Psychology Boston University Boston, Massachusetts

May 2007

Bachelor of Arts Summa Cum Laude with Distinction in Psychology University of Pennsylvania Philadelphia, PA Major: Psychology

GRANTS, FELLOWSHIPS, AND AWARDS 2014

Leonard Krasner Dissertation Award, Association of Behavioral and Cognitive Therapies (Award amount: $1,000)

2013  

Clara Mayo Memorial Fellowship (Award amount: $4,800)

 

98

2013

Boston University Graduate Writing Fellowship

2012

John and Geraldine Weil Dissertation Award (Award amount: $500)

2007

Morris Viteles Award for Outstanding Research in Psychology, University of Pennsylvania (Award amount: $500)

2007

Distinction in Psychology, University of Pennsylvania

2006

Psi Chi Summer Research Grant (Award amount: $3,500)

RESEARCH EXPERIENCE Doctoral Research Assistant Translational Research Program Boston University Advisor: Michael Otto, Ph.D.

 

2009 - 2014

•

“Exposure, D-cycloserine Enhancement, and Genetic Modulators in Panic Disorder” (PI: Michael Otto, Ph.D., R01 MH081116) o Provided individual therapy and outcome assessment for individuals diagnosed with panic disorder in 5-session cognitive behavioral therapy protocol examining the efficacy of D-cycloserine augmentation compared with placebo o Managed study recruitment and screening procedures and IRB review process

•

“Cognitive Remediation with D-cycloserine” (PI: Eden Evins, M.D., R21 DA030808) o Conducted cue exposure therapy sessions with recently quit smokers as part of a federally-funded study of the efficacy of D-cycloserine enhancement of cue exposure therapy to prevent relapse to smoking o Conducted data management and analysis and manuscript preparation

•

“Stress, Distress Intolerance, and Drug Dependence” (PI: Michael Otto, PhD., R01 DA17904) o Served as independent evaluator for treatment development study conducted within two community methadone maintenance clinics

•

Responsibilities also include development and management, including participant selection, protocol management, data quality control and/or analysis, and personnel management for projects including: o A randomized trial examining the efficacy of cognitive control training (CCT) as adjunctive treatment to a brief behavioral activation therapy protocol in the treatment of depression

 

99 o A randomized experimental trial examining the influence of a brief, Internet-based attentional bias modification program on cigarette craving in smokers o A meta-analysis (in preparation) of outcomes of psychosocial interventions for suicidality o Prospective survey studies of trait impulsivity as a moderator between the intention-behavior relationship for substance use, exercise, and other behavioral goal o An experimental study examining the effect of cognitive control training (CCT) on the relationship between repeated checking and memory distrust in healthy and depressed samples

Research Coordinator 2007 - 2009 Center for Anxiety and Traumatic Stress Disorders Massachusetts General Hospital, Boston, MA Supervisors: Mark Pollack, M.D. and Naomi Simon, M.D., M.Sc. •

Managed federally and industry-sponsored research studies of the etiology and treatment of mood and anxiety disorders. Collected and processed blood and saliva samples for studies of the link between psychiatric illness and biological markers of stress. Prepared grant proposals, IRB submissions, posters, and manuscripts. Managed centralized clinic database and conducted data analyses.

CLINICAL EXPERIENCE Predoctoral Psychology Intern September 2014 - present The Institute of Living/Hartford Hospital Supervisor: Elizabeth Moore, Ph.D. • Dialectical Behavior Therapy Intensive Outpatient Program o Co-led group-based DBT skills training o Conducted individual skills coaching and intake interviews • Anxiety Disorders Center/Center for CBT o Conduct CBT with individual adult patients with anxiety and mood disorders in both weekly outpatient and accelerated treatment formats o Co-lead CBT group within Young Adult Services Intensive Outpatient Program o Assist in pilot study of combined CBT/physical therapy group treatment for adolescents with chronic pain • Adult Inpatient Psychiatric Services o Serve as case manager responsible for intake assessment and discharge planning o Conduct brief individual therapy and lead daily CBT groups o Work in close collaboration with interdisciplinary treatment team and coordinate care between inpatient and outpatient providers  

 

100

Study Therapist Translational Research Program, Boston University Supervisor: Michael Otto, Ph.D. • •

•

Conducted brief behavioral activation therapy with 34 adults with major depressive disorder as part of dissertation research Provided ultra-brief CBT for individuals diagnosed with panic disorder in 5session protocol examining the efficacy of D-cycloserine augmentation compared with placebo Conducted study outcome assessment (Panic Disorder Severity Scale, Montgomery-Asberg Depression Rating Scale, Hamilton Anxiety Inventory)

Staff Therapist Center for Anxiety and Related Disorders, Boston University Supervisors: Michael Otto, Ph.D., Heather Murray, Ph.D. • •

2010 - 2014

2010 - 2013

Provided evidence-based treatments to adult individuals diagnosed with a range mood and anxiety disorders Co-led two 12-week cognitive-behavioral therapy groups for individuals diagnosed with social anxiety disorder

Diagnostic Interviewer 2010 - 2013 Center for Anxiety and Related Disorders, Boston University Supervisors: Michael Otto, Ph.D., Heather Murray, Ph.D., Timothy Brown, Ph.D. •

Conducted semi-structured intake assessments with individuals seeking treatment for mood and anxiety disorders using the ADIS-IV-L and UCLA Life Stress Interview

Study Therapist Center for Addiction Medicine, Massachusetts General Hospital Supervisor: Michael Otto, Ph.D. •

Conducted cue exposure therapy sessions with recently quit smokers as part of a federally-funded study of the efficacy of D-cycloserine enhancement of cue exposure therapy to prevent relapse to smoking

Practicum Student Freedom Trail Clinic, North Suffolk Mental Health, Boston MA Supervisor: Corinne Cather, Ph.D. •

 

2011 - 2012

2011 - 2012

Provided therapy to individuals diagnosed with schizophrenia and a variety of comorbid conditions. Interventions included psychoeducation, social skills

 

101

•

training, cognitive restructuring, and exposure therapy. Served as group therapist in CBT-based smoking cessation and relapse prevention research protocol for individuals with schizophrenia or bipolar disorder

Practicum Student 2010 - 2011 Psychological Services Center, Boston University Supervisors: Lisa Smith, Ph.D., Rosemary Toomey, Ph.D., Leah Squires, M.A. • •

Provided CBT to individuals diagnosed with mood and anxiety disorders Conducted structured clinical interviews and neuropsychological assessments (WAIS-IV, WMS-IV, Woodcock-Johnson III, etc.)

Independent Evaluator Bay Cove Human Services and Habit-OPCO Methadone Clinics Boston, MA Supervisor: Michael Otto, Ph.D. • •

2009 - 2011

Served as blind rater for treatment development study conducted within two methadone maintenance clinics Conducted pre-, mid-, and post- treatment assessments using the SCID, Addiction Severity Index, Risk Behavior Survey, Hamilton Anxiety Inventory, and the Montgomery-Asberg Depression Rating Scale

TEACHING EXPERIENCE Co-Instructor Cognitive Behavioral Therapy Course Anxiety Disorder Center, Institute of Living Course Instructor Abnormal Psychology Boston University High School Programs Instructor of Record “Perspectives on Mental Illness” Writing Seminar CAS Writing Program, Boston University

Fall 2014

Summer 2014

Fall 2013 & Spring 2014

Teaching Fellow Developmental Psychology Boston University

Spring 2013

Guest Lecturer Psychopharmacology, Developmental Psychology,

2011 - 2013

 

 

102

Abnormal Psychology Boston University Lab Supervisor Translational Research Program, Boston University

2009 - 2014

PROFESSIONAL MEMBERSHIPS Student Member, Association for Behavioral and Cognitive Therapies (ABCT) Student Member, ABCT Behavioral Medicine and Integrated Primary Care SIG Student Member, ABCT Neurocognitive Therapies/Translational Research SIG Student Member, Association for Psychological Science Student Member, Division 12 of the American Psychological Association, Society of Clinical Psychology AD HOC REVIEWER Addictive Behaviors Mental Health and Physical Activity Nicotine and Tobacco Research Neuropsychology MANUSCRIPTS UNDER REVIEW Diefenbach, G.J. Wootton, B.M., Bragdon, L.B., Moshier, S.J., & Tolin, D.F. (2014). Treatment outcome and predictors of Internet guided self-help for obsessivecompulsive disorder. Manuscript submitted for publication. Yamada, T., Moshier, S.J., Otto, M.W. (2014). No site unseen: Predictors of the failure to control recreational Internet use among young adults. Manuscript submitted for publication. PEER-REVIEWED PUBLICATIONS Moshier, S.J., Molokotos, E.K., Stein, A.T., & Otto, M.W. (in press). Assessing the effects of depressed mood and cognitive control training on memory confidence and accuracy following repeated checking. International Journal of Cognitive Therapy. Moshier, S.J., Landau, A.J., Hearon, B.A., Stein, A.T., Greathouse, L., Smits, J.A.J., & Otto, M.W. (in press). The development of a novel measure to assess motives for compensatory eating in response to exercise – the CEMQ. Behavioral Medicine. Otto, M.W., Moshier, S.J., Gordon, D., Simon, N.M., Pollack, M.H., & Orr, S.P. (2014).  

 

103 De Novo fear conditioning across diagnostic groups in the affective disorders: Evidence for learning impairments. Behavior Therapy, 45 (5), 619-629.

Hearon, B.A., Utschig, A.C., Smits, J.A.J, Moshier, S.J., & Otto, M.W. (2013). The role of anxiety sensitivity and eating expectancy in maladaptive eating behavior. Cognitive Therapy and Research, 37, 923-933. Moshier, S.J., Ewen, M., & Otto, M.W. (2013). Impulsivity as a moderator of the intention-behavior relationship for illicit drug use in patients undergoing treatment. Addictive Behaviors, 38, 1651-1655. Moshier, S.J., Hearon, B.A., Calkins, A.C., Szuhany, K.L., Utschig, A.C., Smits, J.A.J., & Otto, M.W. (2013). Clarifying the link between distress intolerance and exercise: Elevated anxiety sensitivity predicts less vigorous exercise. Cognitive Therapy and Research, 37, 476-482. Moshier, S.J., McHugh, R.K., Calkins, A.W., Hearon, B.A., Rosellini, A.J., Weitzman, M.L., & Otto, M.W. (2012). The role of perceived belongingness to a drug subculture among opioid-dependent patients. Psychology of Addictive Behaviors, 26, 812-820. Deckersbach, T., Moshier, S.J., Tuschen-Caffier, B., & Otto, M.W. (2011). Memory dysfunction in panic disorder: An investigation of the role of chronic benzodiazepine use. Depression and Anxiety, 28, 999-1007. Pollack, M. H., Hoge, E.A., Worthington, J. J., Moshier, S. J., Wechsler, R., Brandes, M., & Simon, N. M. (2011). Eszopiclone for the treatment of posttraumatic stress disorder and associated insomnia. Journal of Clinical Psychiatry. 72, 892–897. Simon, N. M., Worthington, J. J., Moshier, S. J., Marks, E. H., Hoge, E.A., Brandes, M., Delong, H., & Pollack, M. H. (2010). Duloxetine for the treatment of generalized social anxiety disorder: a preliminary randomized trial of increased dose to optimize response. CNS Spectrums, 15, 436-443. Hoge, E., Brandstetter, K., Moshier, S.J., Pollack, M.H., Wong, K., & Simon, N.M. (2009). Broad spectrum of cytokine abnormalities in panic disorder and posttraumatic stress disorder. Depression and Anxiety, 26, 447-455. Hunt, M.G., Moshier, S. J., & Milonova, M. (2009). Brief cognitive-behavioral Internet therapy for irritable bowel syndrome. Behaviour Research and Therapy, 47, 797802. Hunt, M.G., Milonova, M., & Moshier, S.J. (2009). Catastrophizing the consequences of  

 

104 GI symptoms in irritable bowel syndrome. Journal of Cognitive Psychotherapy, 23, 160-173.

Marques, L., Kaufman, R.E, LeBeau, R.T., Moshier, S.J., Otto, M.W., Pollack, M.H., & Simon, N. M. (2009). A comparison of emotional approach coping (EAC) between individuals with anxiety disorders and healthy controls. CNS Neuroscience & Therapeutics, 15, 100-106. Ostacher, M.O., LeBeau, R.T., Nierenberg, A., Lund, H.G., Moshier, S.J., Sachs, G.S., & Simon, N.S. (2009). Cigarette smoking is associated with suicidality in bipolar disorder. Bipolar Disorders, 11, 766-771. Simon, N.M., Otto, M.W., Worthington, J.J., Hoge, E., Thompson, E., LeBeau, R.T., Moshier, S.J., Zalta, A., & Pollack, M.H. (2009). Next step treatment for panic disorder refractory to initial pharmacotherapy. Journal of Clinical Psychiatry, 70, 1563-1570. CHAPTERS Moshier, S.J., Calkins, A.W., Kredlow, M.A., & Otto, M.W. (in press). Neurocognitive perspectives on anxiety disorders and their treatment. In J. Mohlman, T. Deckersbach, A. Weissman (Eds.). From symptom to synapse: A neurocognitive perspective on clinical psychology. Moshier S. J. & Otto, M. W. (2013). Bipolar disorder (pp. 1189-1214). In S. G. Hofmann (Ed.). The Wiley Handbook of Cognitive Behavioral Therapy. UK: Wiley. PRESENTATIONS Moshier, S. J., Yamada, T., & Otto, M.W. (2014, November). Distress Intolerance Mediates the Depression-Exercise Relationship Among Young Adults. Poster presented at the 48th annual convention for the Association of Behavioral and Cognitive Therapies, Philadelphia, PA. Yamada, T., Moshier, S. J., Otto, M.W. (2014, November). Predictors of the failure to control recreational Internet use among young adults. Poster presented at the 48th annual convention for the Association of Behavioral and Cognitive Therapies, Philadelphia, PA Moshier, S. J., & Otto, M.W. (2013, November). Cognitive control training as an adjunct to brief behavioral activation therapy in the treatment of depression. In A. Fang & A. W. Calkins (Chairs), Brain matters! The Importance of neuroscience in depression and anxiety interventions. Symposium conducted at the 47th annual  

 

105 convention for the Association of Behavioral and Cognitive Therapies, Nashville, TN.

Moshier, S.J., Landau, A.J., Stein, A.T., & Otto, M.W. (2013, November). The role of impulsivity in the intention-behavior gap: The example of exercise. Poster presented at the 47th annual convention for the Association of Behavioral and Cognitive Therapies, Nashville, TN. Ward, A.C., Glenn, C.R., Nock, M., Moshier, S.J., & Otto, M.W. (2013, November). A meta-analytic review of the efficacy of interventions for suicidal behaviors. Poster presented at the 47th annual convention for the Association of Behavioral and Cognitive Therapies, Nashville, TN. Stein, A.T., Landau, A., Moshier, S. J., & Otto, M.W. (2013, November). Evaluating motives for compensatory eating in response to exercise: Factor analysis and personality correlates. In S.J. Moshier (Chair), Contextual factors in maladaptive eating: novel research findings and implications for treatment. Symposium presented at the 47th annual convention for the Association of Behavioral and Cognitive Therapies, Nashville, TN. Moshier, S.J., Ewen, M., & Otto, M.W. (2012, November). Impulsivity as a moderator of the intention-behavior relationship for illicit drug use in patients undergoing treatment. Poster presented at the 46th annual convention for the Association of Behavioral and Cognitive Therapies, National Harbor, MD. Szuhany, K.S., Hearon, B.A., Utschig, A., Moshier, S.J., Smits, J.A., & Otto, M.W. (2012, November). Anxiety sensitivity and exercise avoidance: Current status and new findings. In E. McGlinchey (Chair), Physical activity and exercise in psychiatric disorders. Symposium presented at the 46th annual convention for the Association of Behavioral and Cognitive Therapies, National Harbor, MD. Moshier, S.J., Utschig, A., & Otto, M.W. (2011, November). Exercise avoidance and anxiety sensitivity in cardiovascular disease. In B.A. Hearon (Chair), Expanding the application of anxiety sensitivity to health Behaviors: Enhancing treatment across diverse domains. Symposium presented at the 45th annual convention for the Association of Behavioral and Cognitive Therapies, Toronto, ON. Moshier, S.J., McHugh, R.K., Calkins, A.W., Hearon, B.A., Rosellini, A.J., Weitzman, M.L, & Otto, M.W. (2010, November). A psychometric evaluation of the belongingness to drug culture questionnaire. Poster presented at the 44th annual convention for the Association of Behavioral and Cognitive Therapies, San Francisco, CA. Hoge, E., Brandstetter, K.A., Moshier, S.J., Pollack, M.H., Wong, K., & Simon, N.M.  

 

106 (2008, March). A detailed assessment of cytokine abnormalities in panic disorder and posttraumatic stress disorder. Poster presented at the 28th annual meeting of the Anxiety Disorders Association of America, Santa Ana Pueblo, NM.

Pollack, M.H., Moshier, S.J., Hoge, E., Worthington, J., Brandes, M., & Simon, N.M. (2008, March). Effect of treatment for insomnia with eszopiclone in PTSD. Poster presented the 28th annual meeting of the Anxiety Disorders Association of America, Santa Ana Pueblo, NM. Moshier, S.J., Kaufman, R.E., Worthington, J.W., Hoge, E., Pollack, M.H., & Simon, N.M. (2008, November). Subsyndromal mood spectrum symptoms linked to greater stress and poorer quality of life in social anxiety disorder. Poster presented at 42nd annual convention for the Association of Behavioral and Cognitive Therapies, Orlando, FL. Hunt, M.G., Moshier, S.J., & Milonova, M. (2008, November). Internet therapy for IBS - Three month follow-up data. Poster presented at 42nd annual convention for the Association of Behavioral and Cognitive Therapies, Orlando, FL. Hunt, M.G., Milonova, M., & Moshier, S.J. (2006, November). Catastrophic implications of GI symptoms in irritable bowel syndrome. Poster presented at the 40th annual convention for the Association of Behavioral and Cognitive Therapies, Chicago, IL.